Subscribe now

This week, I sat down with Steve van Zutphen, former CEO of Magpie Polymers, acquired by Italmatch Chemicals in 2017, and current CEO of Swan-H.

Drawing on his experience building and leading Magpie Polymers, Steve joined us for a practical conversation about what it really takes to build a business around an advanced material.

We explored why a scientifically credible application can still be the wrong market, how technical founders can improve customer discovery, why industrial sales is fundamentally a process of reducing risk, and how pricing, pilots, and customer collaboration shape the path from laboratory validation to market adoption.

Key takeaways from the episode:

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

Find a market that is willing to pay

The first key takeaway from the conversation is that a technology can solve a real problem and still struggle to become a viable business.

At first, the polymer seemed naturally suited to wastewater treatment and pollution control. It could selectively capture metals, and the environmental need was easy to understand.

The team met potential customers, attended trade shows, received samples, and ran tests. Companies were interested enough to engage, but those conversations did not develop into a clear path to paying customers.

This is an important distinction for advanced materials founders.

A customer may be willing to discuss a technology, test it, or provide feedback without being willing to pay enough.

The problem was not the technology, but the economics of the application. By entering this market, the startup was effectively asking customers to pay more for a better way of managing an existing expense.

That did not mean the market had no value. It meant that the commercial dynamics worked against premium pricing.

For an early-stage chemical company, this is a difficult position.

The company is still producing limited volumes, and manufacturing costs are relatively high. It’s rarely ready to compete through large volumes and low prices.

It needs applications where small quantities of product can create enough value to support healthy margins. This is where product design and market positioning become critical.

Moving from waste management to value recovery

The opportunity became more attractive when the company shifted its focus from removing unwanted metals to recovering valuable ones.

The underlying technology did not need to change substantially. Its core function remained the same. What changed was the customer’s reason to buy.

In a pollution-control application, the polymer helped manage a liability. In precious-metal recovery, it helped recover an asset.

Gold, platinum, and palladium leaving a production process are not simply contaminants. They are materials with a clear value. Recovering them can create a measurable economic return.

This made the commercial proposition much easier to understand.

The customer was no longer being asked to pay primarily for the management of a waste-related cost. The customer was being offered a way to recover value that was already being lost.

Precious-metal recovery also matched the realities of an early-stage materials company much better.

The startup did not need to sell enormous quantities of polymer for the product to matter. Small volumes of a high-value specialty material could support an economically attractive application.

This supported a low-volume, high-value business model.

The broader lesson is that the first market should not be chosen only because the technology can work there or because the need appears important. It should be chosen because the economics fit the company’s current capabilities.

Let the customer talk

Another interesting insight from the conversation was how the team conducted its early customer meetings.

If the founder speaks for most of the meeting, the company may leave without learning the most important things: how the customer’s process works, where the real problems are, and whether solving them is important enough to justify testing a new technology.

The arrival of an experienced salesperson changed the way these meetings were conducted.

The salesperson did not need to understand the chemistry better than the scientists. The value came from knowing how to guide the conversation.

Instead of explaining the product at length, the team began asking simple questions:

• How does the production process work today?

• Where are materials being lost?

• Which steps are expensive or unreliable?

• What has the customer already tried?

These questions helped uncover problems that would never have emerged from a traditional technical presentation.

A productive meeting is therefore not one in which the founder explains everything perfectly. It is one in which the customer shares enough information for the startup to understand whether there is a real opportunity.

Let the problem come before the product

The order of the conversation matters. In the case discussed, the sales team created space for the customer to describe where value was being lost in the process.

Once the problem was clearly understood, the technology could be introduced as a potential solution.

This approach changed the direction of the sale.

In other words, the startup is not asking the customer to test an interesting new material without context. It is proposing a possible solution to a problem the customer has already described.

This makes the next step easier. A pilot is no longer a generic experiment. It becomes a practical test of whether the technology can reduce a specific and measurable loss.

However, as discussed in the conversation, scientific expertise does not automatically make someone effective in sales.

Scientists are trained to explain ideas clearly, and demonstrate technical depth. Customer discovery requires a different skill set: listening, asking useful questions, and waiting before proposing a solution.

For an advanced materials startup, adding commercial experience can therefore be as valuable as adding technical expertise.

Customer adoption means managing risks

Industrial customers do not evaluate a new material only by asking whether it performs better. They also ask what could go wrong if they adopt it.

From the founder’s perspective, the value of the product may seem obvious. The material can reduce costs, recover more value, improve efficiency, or solve a meaningful technical problem.

But the customer sees the decision from a different angle.

An industrial process may not be perfect, but it is already running. Introducing a new material means changing part of an existing system, and that creates risk.

The downside often matters more than the upside

In the discussion, an example helped illustrate the dynamics of adoption in industrial B2B markets. A factory manager may be interested in a solution that reduces costs by 10%. However, if the same solution causes the production line to stop for two days, the loss can be far greater than the expected savings.

This is why the potential benefit and the potential risk do not carry the same weight.

For Deep Tech founders, this helps explain why a technically superior product can still face resistance.

The customer may understand the technology perfectly well. The hesitation may come from the fact that the operational risk has not yet been reduced enough.

This is why it is so important for the startup to understand how the material behaves inside the customer’s full process.

The customer wants to know that solving one problem will not create another, potentially larger one elsewhere in the system. The more the startup understands the customer’s process, the more credible its case becomes.

De-risking comes before price

One of the most useful points from the conversation is that, once the solution has been sufficiently de-risked, price often becomes less central in technical B2B sales.

Customers may spend relatively little time negotiating the final price compared with the time required to evaluate whether the solution works reliably without disrupting the process.

This does not mean price is irrelevant, but before discussing small differences in cost, the startup has to build confidence.

The customer needs evidence that the material performs consistently, fits into the existing process, and does not introduce risks that are difficult to manage.

Once that confidence exists, the commercial conversation becomes much easier.

The broader lesson is that industrial adoption is rarely driven by performance alone. It is driven by the combination of measurable value and acceptable risk

Pricing and business models

As we’ve discussed many times on this channel, for a Deep Tech startup, the question is not only how much to charge. It is also what the company should sell.

In advanced materials, this is particularly important.

The offering may be the chemical itself, a machine that doses it, a packaged system, or a service built around the result. Each option creates a different level of operational complexity, resource requirements, and customer support.

The most attractive model on paper is not always the most practical one for an early-stage company.

Selling a product is often a practical choice.

The customer can receive the chemical and introduce it into the process directly. In some cases, the startup may also provide a machine that doses the material.

This can be useful during early adoption because it gives the company more control over how the product is used and allows it to work more closely with the customer.

The machine does not necessarily need to be built internally. A specialist supplier can manufacture it, while the startup focuses on the chemistry that makes the system valuable.

The equipment supports adoption, but the chemistry creates the margin

The conversation drew a useful distinction between the economics of the machine and the economics of the material.

Industrial equipment can often be sold with a relatively conventional margin. The machine helps the customer adopt the technology and gives the startup greater control over the application.

But it is unlikely to be the main source of long-term value. The chemical is different.

Its price has to reflect more than the cost of manufacturing a kilogram of product. Years of R&D, testing, application development, and customer work sit behind that material. A simple cost-plus approach can therefore be misleading.

Moreover, at the beginning, founders may focus too heavily on maximizing the first sale.

For an early-stage startup, acquiring a first and then a second customer can matter more than maximizing the revenue generated by a single account.

That matters because early revenue is not just cash. It is proof that the commercial model may be repeatable.

The broader lesson is that pricing decisions should support company building, not just short-term revenue. They should also help the company reach the next customer, prove repeatability, and move toward a business that can scale.

From the lab to the customer site

Product design does not always have to happen entirely inside the startup. In the case discussed, much of the learning about how to use the material happened directly at customer sites.

This created a practical division of tasks.

The startup focused on mastering the production of the polymer. The customer helped validate how the material could be used inside a real industrial process.

This allowed the startup to retain control over material production while developing the application in the customer’s operating environment.

On-site development

The process began with small lab tests. Customers sent samples from their operations, and the team used them to see whether the polymer worked with real process materials.

If the results were promising, testing moved to the customer’s facility.

The team brought small mobile pilot units to the site. Testing could then move to larger or custom-built systems to evaluate the product at a greater scale.

This created a step-by-step path from lab testing to industrial use.

• The lab showed that the chemistry worked.

• On-site testing showed that it worked under the customer’s actual operating conditions.

The larger system helped the customer determine whether the technology could work at a larger scale.

Moreover, working at the customer’s site meant sharing part of the learning process.

The customer became familiar with the product and developed direct experience with how it worked and which risks needed to be managed.

The co-development phase therefore did more than validate performance. It reduced uncertainty and reduced adoption risk.

By the time the customer considered a larger-scale purchase, the decision was no longer based only on the startup’s claims. The product had already been tested inside the customer’s facility.

Deciding whether to charge for early trials

The conversation also highlighted a difficult strategic decision: whether and how much to charge for early trials.

Charging for early testing or co-development work can help the startup cover its costs. But insisting on full payment may slow the project and delay progress with the customer.

At the same time, the startup has a strong interest in entering the customer’s facility and moving the project forward, even if that sometimes means progressing without charging the full cost.

A practical solution could be to share the investment.

The company can explain what the test will cost and ask the customer to contribute. The payment may cover part of the project, for instance, or later be converted into a discount on a full-scale purchase.

The exact structure can vary. What matters is that both sides have something invested in the process.

The startup needs to protect its limited capital, but the long-term objective is not to build a business around fees from early-stage collaborations. It is to reach full-scale adoption and recurring sales.

Learning to filter the wrong opportunities

The final takeaway from the conversation is that not every customer project deserves the same attention. A CEO must therefore stay focused on the right opportunities.

Early-stage startups can easily become excited by any sign of interest. A customer mentions a technical problem, and the team immediately begins designing experiments around it.

But some projects are more valuable than others.

One customer may have a real problem that needs solving. Another may offer a project only because the startup has repeatedly asked for one. A third may present a challenge that has already been tested many times and may not be realistically solvable.

The first meeting is not enough.

Founders should avoid becoming overly optimistic, let the discussion settle, and see whether the customer still has the same problem the following week.

The broader lesson is that advanced materials startups need to protect their time and focus as carefully as their capital. A promising conversation is not yet a commercial opportunity. The strongest opportunities are those in which the customer has a genuine problem and the startup can realistically solve it.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape effectively.

The Scenarionist Premium is designed to make you a better Deep Tech Founder, Investor, and Operator. Premium members gain exclusive access to unique insights, analysis, and masterclasses with the wisdom of the world’s leading Deep Tech thought leaders. Invest in yourself, and upgrade today!

Subscribe now

Deep Tech venture design often doesn’t start from an unmet customer need, but from a technology with enabling potential.

A new material. A novel manufacturing process. A scientific discovery. A proprietary algorithm. A new approach to energy conversion, sensing, computation, biology, or industrial production.

Something becomes technically possible that was previously impossible.

This is the source of Deep Tech’s extraordinary potential. It is also the source of its strategic challenge.

A new technology rarely points toward a single specific product, customer, or market. It often enables multiple possibilities.

It may improve performance in one application, lower costs in another, shorten a critical process somewhere else, or make an entirely new outcome possible in a market the founding team had never considered.

That broad spectrum is often presented as an advantage.

The company accumulates use cases, customer conversations, and technically plausible directions. Each one appears to confirm that the opportunity is larger than initially imagined.

However, a broad optionality creates value only when the company knows what to pursue now, what to preserve for later, and what conditions must be true before the next move becomes rational.

Multiple applications may point to scale and larger markets. They can also fragment attention, accelerate cash burn, and impose significant opportunity costs.

The question is not simply where the technology can be used. It is where it can create an advantage powerful enough to change a customer’s decision and durable enough to support lasting value capture.

Subscribe now

The logic of this piece

So, in Deep Tech, what matters most when building a strong competitive advantage?

Deep Tech is industry-specific and product-specific. The dynamics of a semiconductor company are not those of a mining technology, a medical device, or an advanced materials venture.

Still, some strategic considerations travel across sectors. In moments of company creation, market entry, or strategic pivot.

This was the starting point that inspired this piece.

The answer does not lie in technology alone. It may depend on the business model, distribution agreements, access to customers, control of critical resources, or the company’s position within the value chain.

Drawing on the experiences and conversations collected over time, as well as the lens developed through direct engagement with people operating in the field, I have organized the piece to provide some curated food for thought.

The objective is to move from broad technological optionality to the strongest competitive edge around which a company can build a coherent and durable advantage.

Each section includes practical questions for founders and investors working through market selection, product direction, strategic pivots, and venture design.

From here onward, only our Premium Members can keep reading.

Read more

Welcome to Edition N^o 115 of Deep Tech Briefing.

The Scenarionist’s weekly intelligence layer on what you need to know, understand, and track to win in deep tech: company milestones, market shifts, and macro forces reshaping outcomes, competitive position, capital allocation, and critical decisions.

In this edition, the Big Idea asks what happens when a critical industrial input is no longer purchased shipment by shipment, but produced inside the customer’s own operation.

The immediate case begins with a single unit operation inside rare-earth recycling. The larger question reaches much further: when a supplier moves onsite, does the customer gain independence, or enter a deeper and more durable relationship with the technology provider?

Subscribe now

Beyond it, this edition moves across sovereign radar intelligence, physics AI for engineering, AI-native industrial robotics, second-life batteries, sodium-ion storage, autonomous freight, commercial e-fuels, photonic-chip manufacturing, compact fusion, distributed quantum computing, and deep-sea geothermal.

The market-shaping layer follows the infrastructure around those technologies: public guarantees, cyber standards, faster defence permitting, classified market access, public demand for specialised AI hardware, and new research and technology-transfer pathways.

Finally, the edition closes with 10 key data points to know — from a valuation above €10 billion and 465 Wh/kg at stack level to 100 repurposed EV battery packs entering an industrial storage system.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Subscribe now

When On-Site Recovery Replaces Consumables, Who Captures the Value?

Most industrial systems are built as lines. Materials enter, processes transform them, products leave, and waste moves elsewhere.

Each stage depends on a network of external relationships: energy providers, chemical suppliers, logistics operators, water systems and waste contractors.

That arrangement is often efficient. Specialization allows each participant to do one thing well, while the plant avoids having to reproduce every capability internally.

But what happens when one of those relationships becomes unavailable, unreliable or simply no longer economical?

This is what makes the collaboration between

Read more

Subscribe now

This week, I sat down with Anil Achyuta, Partner at Energy Impact Partners, for a practical look at the battery sector from the perspective of founders building across the value chain — from materials, cells, manufacturing processes, and packs to grid storage, EVs, and robotics.

We explore why battery startups are difficult to build in a capital-efficient way, how demand is shifting across end markets, what makes cell manufacturing so industrially demanding, and why a promising material is rarely enough unless it can survive scale-up, qualification, and real customer adoption.

Key takeaways from the episode:

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

The Battery Market Is Being Rewritten by Demand, Not Chemistry Alone

The battery market is shaped by chemistry, materials science, manufacturing capability, and electrochemical performance. But before any of those dimensions can become investable, there has to be a strong enough market pull. Without demand, even a strong technical breakthrough can remain trapped at the level of promising experimentation.

Batteries sit inside larger industrial systems. They are an input into electric vehicles, grid storage, robotics, consumer electronics, defense systems, and many other applications.

This matters because different markets move at different speeds.

They have different qualification requirements, different cost sensitivities, different safety expectations, and different levels of tolerance for new technology. A breakthrough that looks compelling in one application may be irrelevant, too expensive, or too difficult to qualify in another.

Grid storage

One of the clearest sources of demand today is grid storage.

The reason is structural. Power systems are under pressure. AI-related energy demand is growing. Grid upgrades are becoming more important.

Countries and regions around the world are building large amounts of renewable energy, and renewables create a basic systems problem: generation does not always arrive when demand needs it.

Batteries become part of the infrastructure that makes renewable power more useful, and more compatible with the needs of a modern electricity system.

For founders, this creates a different kind of market environment.

However, utilities and grid operators can be slow-moving customers. They care deeply about reliability, safety, and long-term performance. But the underlying demand is real, and that matters.

The EV market has become more complex

Electric vehicles remain one of the largest battery demand pools in the world.

The reason is obvious but important: cars require a lot of battery mass. A vehicle can carry hundreds of kilograms of batteries, which means automotive demand creates enormous volume.

When the EV market accelerates, it can pull the entire battery value chain with it, from raw materials to cells, packs, battery management systems, and manufacturing equipment.

But the EV market has also become more complex. Many automakers have explored verticalizing battery manufacturing by bringing it in-house, with the goal of reducing costs and making EVs more affordable at scale. However, a changing policy context and market environment have made some automotive players more cautious.

Robotics is growing fast

Robotics is one of the most exciting emerging demand areas for batteries.

The category is expanding across many forms: delivery robots, sidewalk robots, autonomous systems, lawnmower robots, cleaning robots, specialized industrial robots, and humanoid robots.

That makes robotics interesting, but it’s not yet the same as automotive in volume.

Every Battery Market Has Its Own Performance Logic

In batteries, the right performance profile depends on the application. A consumer device, an electric car, a grid storage system, a drone, and a humanoid robot do not need the same thing from a battery.

Founders often think in terms of superior chemistry or superior materials. Customers think in terms of systems performance.

A founder may believe the company has a better anode, a safer electrolyte, a more energy-dense cell, or a lower-cost manufacturing process.

But the customer wants to know what that improvement means inside the actual product.

• Does the car get more range?

• Does the grid asset cycle longer?

• Does the robot work for more hours?

• Does the cost fall enough to change adoption?

Technical improvement has to translate into a use-case advantage.

Energy density, cycles, safety, and cost

When it comes to performance, electric vehicles offer a useful example.

One of the most important metrics is energy density.

At the simplest level, energy density determines how much energy can be stored per unit of weight or volume. In a car, this connects directly to range. The more usable energy a vehicle can carry without adding too much weight, the more attractive the vehicle can become.

But energy density is only one part of the equation.

Cyclability matters because batteries degrade. Founders and customers need to understand how many times a battery can be charged and discharged over its useful life.

Safety is another core requirement.

Batteries need to operate safely across different conditions, including high and low temperatures. A battery that offers strong performance but cannot meet safety expectations will struggle to earn adoption.

Power density adds another layer.

Some applications need to discharge power quickly, while others may prioritize different performance characteristics. Temperature performance matters as well, especially for systems operating in demanding environments.

Then there is cost.

Cost is not a secondary metric. It is one of the constraints that shapes whether a battery technology can become commercially relevant.

A battery can be technically impressive and still commercially weak if it cannot reach the cost structure required by the market.

In cars, for example, battery cost can represent thousands of dollars per vehicle. The broader vehicle cost then builds on top of that foundation. This is why battery cost has strategic consequences far beyond the cell itself.

Cell Manufacturing Is Where the Value Chain Becomes Industrial

The battery value chain begins with raw materials and extends through processing, active materials, electrodes, cells, modules, packs, battery management systems, and final integration into vehicles, grids, robots, or devices.

Each layer has its own technical requirements and commercial constraints.

This is important for founders because a startup may enter the value chain at one narrow point, but the market will eventually judge whether that point works inside the full system.

From lithium to cells to packs

A simplified view of the battery manufacturing process helps clarify why this market is so demanding.

Lithium first has to be extracted, whether from rock or brines. It then has to be processed and refined into usable forms such as lithium carbonate or lithium hydroxide. These materials enter the manufacturing chain and are combined with cathode active materials and other inputs.

The next step is electrode production.

Materials are mixed and coated onto surfaces through processes such as slurry casting. The coated material is then processed to create uniform layers, often through calendaring, which helps produce thin, consistent electrode structures.

The cathode and anode are prepared separately, with anodes commonly involving graphite and, increasingly, graphite-silicon combinations.

Those components are then brought together with separators and electrolyte. The cell goes through formation, where it is electrochemically cycled and prepared for use. Depending on the format, the cell may be rolled, packaged, tabbed, and assembled into its final form.

But the cell is still not the complete product.

Cells are assembled into modules or packs. Packs require electrical connections, mechanical integration, packaging, and a battery management system (BMS). The BMS acts as the intelligence layer that communicates with and coordinates the behavior of the cells.

Only then does the battery become part of the final system.

For a founder working on one material, one process step, or one component, this process creates a hard reality: the innovation has to fit into and perform across every downstream layer.

A gigafactory is not just a bigger lab

The move from laboratory work to manufacturing is not a linear scale-up. It is a different operating environment.

It requires capital, a manufacturing line, repeatable processes, and the ability to produce cells reliably in the same way over and over again.

One rough reference point from the conversation is that a 10 gigawatt-hour battery manufacturing plant in the US can require hundreds of millions of dollars in capital expenditure, before even considering the operating costs required to keep the facility running.

That figure is not meant to be a precise universal rule. The exact number can vary depending on the project, but it captures the scale of the challenge.

Battery manufacturing is capital intensive because the product is physical, the process is complex, and the market requires reliability at scale.

Owning One Breakthrough Is Not Enough

A team may develop a better anode, a new cathode material, a safer electrolyte, or a novel manufacturing process.

At the scientific level, the breakthrough may be real. It may show impressive performance in early testing. It may even look compelling in techno-economic models.

But batteries are systems.

A founder can be world-class at one component and still be forced to prove performance at the cell level. Customers do not buy an anode in isolation. They need confidence that the full cell will behave reliably and consistently inside the broader battery system.

This changes the company-building journey.

A better anode still has to work inside a cell

Take the example of a silicon anode.

A founder may have a strong technical thesis and compelling data showing that the anode improves performance. But that anode still has to fit into a cell. It has to interact with the cathode, separator, electrolyte, and manufacturing line.

The customer wants to know whether the new material behaves predictably in the full system.

• Does the cell perform reliably?

• Can it be manufactured reliably in the same way every time?

• Does the improvement still hold once it is tested inside the complete battery system?

These questions are the reality of adoption. Every improvement has to be validated inside the system that will carry it to market.

The founder becomes a cell manufacturer without wanting to

This creates one of the most difficult capital efficiency problems in battery startups.

A founder may begin as a materials company. The company’s expertise may be concentrated in one part of the stack.

But to prove that the material works, the founder may have to build cells. To build credible cells, the company may need a manufacturing line, physical infrastructure, and process know-how.

Before long, a materials company starts behaving like a cell manufacturer.

This may not be the founder’s original strategy. It may not be where the company has the strongest expertise. But the market demands system-level proof, and system-level proof often requires physical infrastructure.

That is why battery startups can burn capital even when they are trying to remain focused.

The problem is that the proof required by the customer sits downstream from the original innovation. If the ecosystem does not provide enough external infrastructure to bridge that gap, the startup has to build more itself.

This is one of the reasons the battery sector is so complex.

A company can have a real breakthrough and still face a financing path that looks long, expensive, and dependent on manufacturing proof, customer qualification, and eventual volume.

The missing contract manufacturing ecosystem

One important structural issue discussed is the lack of a strong contract manufacturing ecosystem. A useful way to understand that gap is through an interesting parallel with the biotech sector.

In biotech, capital-efficient company formation became more possible because an ecosystem of contract manufacturers developed around the industry.

A startup did not always need to own every piece of manufacturing infrastructure from the beginning. It could rely on external manufacturing partners for parts of the development and production journey.

That does not make biotech easy. It is a difficult market with its own cycles, risks, and capital requirements. But the presence of contract manufacturing infrastructure changed what early-stage companies could attempt without owning everything themselves.

Batteries do not yet have an equivalent layer at sufficient depth.

If there were a broad network of contract manufacturers for batteries, component-level startups could potentially build in a more capital-efficient way. But today, that layer is not deep enough, and OEMs still need to test and qualify the battery themselves.

This is a structural issue.

If a founder could take a new material to a robust network of contract manufacturers, produce credible cells, and generate the data required for customers, the company-building path would look different. It would still be hard, but the burden of infrastructure ownership would be lighter.

In the absence of that layer, founders have to be much more careful.

They need to know early which parts of the manufacturing journey they must own, which parts they can access through partners, and where the proof of value actually has to occur.

Qualification, licensing, and economics

A battery innovation does not become valuable simply because the technology works. It becomes valuable when the market believes the technology can be adopted, manufactured, qualified, and delivered at scale.

That requires more than technical data. It requires a credible path through customers, production, financing, and time.

This is where battery companies often face long timelines. An important milestones may require years of testing, physical infrastructure, and OEM qualification.

For example, qualification cycles for automotive customers can be very long.

Even a strong technology has to prove that it behaves reliably and safely over time. OEMs need confidence that the cell will perform consistently, not only in ideal conditions but across real-world usage. They have to understand reliability, safety, manufacturability, and integration risk.

This process can take years.

The exact timeline will depend on the customer, application, chemistry, and level of integration, but the strategic point is clear: qualification is a central part of the adoption process.

Licensing often appears attractive in battery materials.

In theory, a company can avoid building massive factories by licensing its technology to larger manufacturers. The model can include upfront payments, milestone payments, and royalties over time. It can seem attractive because it appears to reduce the need for the startup to own the full manufacturing stack.

But licensing only works if the licensed technology reaches volume. Royalties depend on production. If volume does not materialize, neither do the future cash flows. Again, the major economic value often depends on adoption at scale.

Founders Should Think Scale First

One of the most important lessons from the episode is that battery founders should think about scale earlier.

This does not mean every startup needs to build a factory immediately. It does not mean every company should raise hundreds of millions of dollars before proving anything. It means the technical roadmap should be designed with manufacturing reality in mind from the beginning.

A founder cannot afford to treat scale as a later question.

Cell size matters

Cell size matters because it signals how close the company is to meaningful proof.

Coin cells and very small cells may be useful for early experimentation, but they are not enough to establish confidence that a battery technology is moving toward real-world relevance. Investors want to see progress toward formats that begin to resemble real-world relevance.

One amp-hour cells are still small. Moving toward five amp-hour cells begins to create a more meaningful signal. The number may vary by application, but the principle is that founders need to show that performance survives when the system becomes more realistic.

Defense, robotics, and grid

For some founders, the smartest path may not begin with the largest market. A startup may need earlier markets that generate stronger near-term pull, faster feedback, or more credible commercial validation.

Dual-use is one possible path.

In defense, customers may care deeply about speed, performance, reliability, and mission capability.

Some of these buyers may move faster than traditional automotive or utility customers. They may also value differentiated energy systems if those systems create a real operational advantage.

But the key is traction. Investors will want to see evidence that real customers are engaging seriously.

Grid storage can also be attractive, especially because demand is structurally strong. But founders have to recognize that utilities and power customers may move slowly.

Reliability, bankability, and long-term economics matter. The market pull is real, but the adoption process still requires patience and proof.

The broader lesson is that founders should not choose markets only by size.

They should choose markets based on where they can prove value, and generate credible customer signal. A smaller early market with urgent demand may be more useful than a massive market that takes years to validate.

The goal is to sequence ambition correctly.

The battery founder’s real job is to reduce industrial risk

To recap, battery startups are difficult because they sit at the intersection of science, manufacturing, capital, and adoption. A founder has to manage all four.

The science has to be real. The manufacturing path has to be credible. The capital plan has to match the proof points. The customer has to care enough to participate in the journey.

This is why the best battery companies are not built around technical optimism alone. They are built around disciplined risk reduction.

Every milestone should answer a specific question.

• Does the material work beyond the lab?

• Does performance survive in larger cells?

• Can the process be repeated?

• Can the cost structure support the target application?

• Will customers test it?

• Will they pay for it?

• Can the company reach the next proof point without consuming more capital than the opportunity can justify?

The strongest founders will show why the technology matters for a specific market, how it fits into the value chain, what it takes to manufacture, where customer pull is strongest, and which proof points will make the company more investable over time.

That is the reality check.

Batteries remain one of the most important sectors in deep tech. The companies that succeed will be the ones that understand this early. They will think beyond the material, beyond the lab result, beyond the spreadsheet, and beyond the first customer conversation. They will build with the full system in mind.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape effectively.

Welcome to Edition N^o 114 of Deep Tech Briefing.

The Scenarionist’s weekly intelligence layer on what you need to know, understand, and track to win in deep tech: company milestones, market shifts, and macro forces reshaping outcomes, competitive position, capital allocation, and critical decisions.

In this edition, the Big Idea is not only about artificial intelligence, and not only about materials.

It is about the measurement layer that may decide whether AI can become useful in the physical industries where matter still behaves in ways that surprise even the best teams.

Economies are built on materials, but materials are not always fully understood. Food, pharmaceuticals, cosmetics, packaging, batteries, coatings, biologics, and specialty chemicals all depend on matter that changes with temperature, humidity, pressure, formulation, purity, processing conditions, storage time, and biological context. Even advanced companies can lose money when a material behaves differently after the capital has already been spent, the line has already been built, or the product has already entered the customer’s workflow.

Subscribe now

Beyond it, this edition moves across satellite manufacturing, counter-drone security, enterprise AI, hydrogen-to-chemicals, greenhouse robotics, solar deployment for data centers, rare earth processing, lithium offtake, battery manufacturing, perovskite supply chains, advanced nuclear, fusion, quantum computing, industrial robotics, commercial space stations, utility pilots, and flood intelligence.

The market-making layer this week moves from space defense and sovereign industrial capacity to critical minerals, AI infrastructure, energy demand, and public-sector procurement.

And finally, the edition closes with a curated selection of 10 key data points to know — from a $2.3 billion valuation to a fleet of 2,000 sidewalk robots.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Subscribe now

How Much Is It Worth to Own Matter Behavior Data?

The old workflow is familiar: formulate, test, run sensory panels, adjust, wait, scale, then discover new behavior inside the factory or distribution chain. It is slow because physical products contain too many linked variables for human iteration alone.

This week, a startup came out of stealth and made that problem worth revisiting. The point is the paradigm shift: physical product development may be moving from measuring what matter is to measuring how matter behaves.

Composition tells one story. Structure tells another. Behavior tells whether a product survives stress, time, interfaces, scale-up, storage, logistics, and use.

So the natural question is: what is the risk of not knowing how matter behaves?

Read more

Subscribe now

This week, I sat down with Jeff Uhrig, a 2x exited Deep Tech founder who is now Principal and Founder of Beech Lab. In this conversation, he reflects on his experience as CEO of Sirrus, a specialty chemicals startup acquired by Nippon Shokubai in 2017.

We explore how the company turned a promising molecule into a commercially relevant platform, and how the team navigated application discovery, customer-led development, scale-up, capital allocation, and ultimately the acquisition.

Key takeaways from the episode:

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

First, Prove You Can Manufacture It

The Sirrus story began with a molecule that had been known for some time but had never been manufactured at a scale and cost that made commercial adoption realistic.

The material could bond surfaces quickly while remaining less reactive than traditional superglues. That made the chemistry interesting, but scientific interest was not enough.

The first real challenge was to develop a process capable of producing the molecule economically and consistently.

The founders came from Loctite, so adhesives were the natural place to begin. Yet when the company looked more closely at the market, adhesives proved to be a demanding entry point.

In adhesive applications, performance requirements were high, and the material often had to bond two very different substrates under specific operating conditions.

The company therefore approached the question with an open mind.

Instead of assuming that the molecule had to succeed in the application originally imagined, the team began asking where it could create value with the lowest technical barrier. That distinction was important.

The first application did not need to represent the full potential of the platform. It needed to give the company a credible way into the market, create early progress, and give investors and commercial partners a reason to continue supporting the technology.

Why a cheaper molecule may cost the customer more

To evaluate possible applications, the team used a simple framework:

In advanced materials, however, cheaper is often the weakest place to begin.

A new molecule may have a lower production cost while still being more expensive for the customer to adopt.

Incumbent materials are often supported by infrastructure that has already been installed. Processes have been optimized, employees have been trained, and customers already understand how the existing solution behaves.

A new material can disrupt that system.

If adoption requires different equipment, changes to the production process, or additional training, the apparent cost advantage can disappear quickly.

A cheaper molecule can become a more expensive solution once the full operational impact is considered.

This pushed Sirrus toward a broader view of value.

The company could not compete only on the price of the material. It had to understand how the material affected the customer’s entire process and where its properties could create a meaningful advantage.

The commercial opportunity would emerge from finding an application where the molecule did more than cost less—where it allowed the customer to operate differently.

From Adhesives to Coatings

The first meaningful commercial shift came when Sirrus stopped looking at the molecule only as an adhesive.

The team had spent time studying where the material could create an advantage, and one property stood out: its very low viscosity.

In practical terms, the material could be sprayed without first being diluted with water or solvent. That opened a different path.

In coatings, solvents are commonly used to make materials easier to apply. They then have to be removed through drying, which adds energy consumption, equipment requirements, and process time.

Because the innovative material could be sprayed without adding water or solvent, it offered the possibility of a solvent-free coating with lower drying requirements.

The molecule also created a strong surface finish through covalent bonding. This made it particularly relevant for clear-coat applications, where the coating must protect the underlying surface from scratching, and chipping.

The opportunity was therefore no longer simply to sell a faster-bonding adhesive. It was to use the same underlying chemistry in an application where its existing properties created a clearer operational advantage.

The first application was discovered through iteration

This direction did not emerge from a single workshop or a design session. It came from repeated technical work, customer conversations, and substantial trial and error.

The team had to understand the material as it existed, while also considering where it could realistically go.

At the same time, it had to understand the customer problem well enough to identify the point where the technology and the application could meet.

That process required constant iteration.

Moreover, the Sirrus team was looking for an application where the material could solve a meaningful problem without requiring the company to redesign the entire platform before entering the market.

The coating opportunity met that standard. It lowered the technical barrier relative to the original adhesive thesis while preserving important performance advantages.

Customer motivation unlocks adoption

Even a strong technical fit, however, was not enough. Sirrus also needed a customer with a clear reason to move the technology through the market.

That became possible through a partnership with a large automotive company that invested in the startup and had a strong interest in reducing both the energy required to cure automotive coatings and the costs associated with drying them.

That motivation changed the adoption dynamic.

Sirrus was not selling directly into an isolated customer relationship. It operated through paint companies and other channel partners. Those suppliers became more interested in the technology because one of their major customers wanted the solution.

This created pull through the channel.

The lesson was that the best application is not always the one that appears most attractive in theory. It is often the one where the customer is sufficiently motivated to help move the technology through a complex value chain.

Selling Economic Outcomes Instead of Chemistry

Once Sirrus identified coatings as a promising application, the next challenge was deciding how to present the technology to the market.

The company could have approached paint manufacturers with a monomer and asked them to determine how it might fit into their formulations.

That would have placed much of the development burden on the customer and made the value of the material harder to see.

Instead, the team built formulation expertise internally.

Paint companies already considered themselves experts in formulation, and each chemistry came with its own technical nuances.

The company therefore positioned itself carefully. Its partners might understand coatings broadly, but Sirrus understood how to formulate its own material better than anyone else.

That capability changed the nature of the conversation.

When the company approached the customer, it did not lead with the molecular structure, covalent bonding, or interpenetrating networks. It presented a formulated coating that could be applied to a base coat and evaluated as a practical solution.

The chemistry remained essential, but it was no longer the center of the commercial message. The customer could see what the material did, how it could be used, and why it might matter inside an existing production process.

Building the price backward from customer value

The same logic shaped Sirrus’s pricing strategy.

Rather than beginning with the cost of producing the monomer and adding a conventional margin, the company started with the value created for the end customer.

In the automotive application, that value came from reducing the energy required to cure coatings and lowering the costs associated with drying equipment.

Sirrus examined how much energy could be saved per vehicle and how much capital expenditure could be avoided or reduced.

Those savings created an economic value for the formulated coating.

From there, the company worked backward through the value chain. It considered what a paint company could charge the customer for the coating and then asked how much of that value Sirrus could reasonably capture as the supplier of the underlying monomer.

This was a more useful approach than pricing the molecule in isolation.

The relevant question was not simply what the material cost to manufacture. It was how much economic value the material enabled once incorporated into the customer’s process.

Sharing enough value to support adoption

Value-based pricing did not mean attempting to capture all the value created.

Based on the conversation, the team worked on the assumption that, to enter the market, the OEM would capture somewhere around 70% of the value created.

That left the startup and its channel partners to divide the remaining benefit.

The pricing decision therefore depended on balance. The company needed to preserve enough value to build an attractive business, while leaving customers with a strong financial reason to adopt the technology.

This also reinforced the importance of understanding the full channel.

Sirrus was not selling only to the final user. It needed to create value for the paint companies that would formulate and supply the coating, while ensuring that the automotive manufacturer still received a compelling share of the savings.

The stronger Sirrus became at translating technical performance into energy savings, capital savings, and value per vehicle, the easier it became to explain where the molecule belonged economically.

The company was no longer asking customers to pay for an interesting chemistry. It was showing them how that chemistry could improve the economics of production—and then pricing the material according to the value it made possible.

Letting Customers Guide the Expansion

For its first commercial application, the team followed a deliberately simple principle: change the underlying technology as little as possible.

The initial objective was to work with the easiest version of the molecule to manufacture and identify applications where it could already create value.

This reduced technical complexity and allowed the startup to move more quickly into collaborations with partners.

That approach mattered because the first market entry was not meant to prove every possible use of the platform. It was meant to create progress with the material the company could already produce reliably.

The team could then learn from real applications before committing additional time and capital to more complex molecular development.

When customer feedback becomes a product roadmap

The second application emerged from a conversation with a supplier serving semiconductor assembly.

The partner saw clear advantages in the existing material. It cured quickly and had low viscosity, both of which made it attractive for the application.

But the customer also identified two important weaknesses: the molecule was too brittle, and its expansion characteristics were poorly suited to repeated temperature changes.

In a semiconductor application, those limitations were critical.

As a chip heats and cools over its operating life, the surrounding materials must expand and contract with it. If the adhesive cannot respond to that thermal history, it can lose adhesion and compromise the application.

The customer’s feedback gave the development team a much more precise target.

The problem was no longer whether the material had interesting properties. It was whether those properties could be adapted to meet the requirements of a specific, valuable use case.

Advancing the material through co-development

Rather than redesigning the chemistry in isolation, the company worked directly with the partner to develop new molecules.

The goal was to preserve the advantages of the original platform while improving flexibility and thermal expansion behavior.

The team produced a range of molecular options based on those hypotheses, and the partner tested them in the end application.

This changed the relationship between technical development and market discovery.

The customer was no longer simply evaluating a finished product. Its application knowledge became part of the development process, helping the team decide which technical improvements mattered most.

That is how the platform expanded: not through a broad search for every possible market, but through a focused collaboration with a partner that understood both the value of the existing material and the limitations that had to be solved next.

From Bench to Scale

For an early-stage materials company, scale-up is not separate from technology development. The two happen at the same time.

As the process moves from bench scale to pilot and demonstration scale, new problems emerge.

Reactions behave differently. Side reactions appear. Material can polymerize inside piping. Steps that looked straightforward in the lab become difficult to control in a larger system.

The team chose to keep that learning inside the company.

Rather than outsourcing the full path from laboratory development to demonstration scale, the company kept that work in its own labs.

That allowed the company to understand the full supply chain and preserve the knowledge created as the process evolved.

Some of that knowledge could be protected through patents. Some was retained as trade secrets. In both cases, the practical experience of solving scale-up problems became part of the company’s intellectual property.

Designing capital efficiency into the process

The underlying manufacturing approach also made the scale-up path more manageable.

The process used continuous-flow chemistry, allowing relatively large volumes to be produced in a small reactor. It also relied on small-diameter piping systems that could be replicated as production requirements increased.

This created a “modular path to scale”, making the manufacturing story easier to understand and the process more capital efficient.

In other words, partners could look at the demonstration-scale system and see how the same basic process could operate at much greater capacity, without assuming that the company still had to solve an entirely new engineering problem.

The real value was held by the people

Over the course of the conversation, it became clear that the company’s equipment mattered, but the people operating it mattered more.

By keeping development and scale-up work inside the organization, the team accumulated the practical knowledge behind the technology: what failed, why it failed, how the process had been improved, and which decisions were essential to making it work.

That knowledge became central when communicating with partners and, later, potential acquirers.

Bottom line: a company is only as valuable as the people who understand what has been built. The more technical learning those people retain, the easier it becomes to explain why the technology is credible, how it can scale, and where its value truly lies.

The cost of expanding too early

Capital efficiency at the process level did not eliminate difficult decisions at the company level.

At the earliest stages of its journey, the company was pursuing opportunities across adhesives, coatings, inks, and sealants. The team expanded to support all 4 areas, reaching roughly 40 employees.

That breadth created excitement, but it also created significant cost before the company had established product-market fit.

When leadership concluded that the business could not succeed across all four markets simultaneously, the organization had to narrow its focus.

Headcount was reduced by more than half, and the company moved from pursuing a wide range of possibilities to concentrating resources on the strongest opportunities.

According to the conversation, over 7 years, the company raised approximately $40 million. Around one quarter was used for capital equipment and installation, while most of the remaining capital went toward people, resources, and advisors.

The difficult lesson was that a broad platform can create too many plausible directions. Without focus, capital can be spent expanding the organization before the market has shown which opportunity deserves to be built.

Exit Strategy

As commercial partnerships advanced, the company reached a decision that many Deep Tech startups eventually face:

• One option was to raise additional capital, build manufacturing capacity, and develop the commercial infrastructure needed to scale independently.

• The other was to allow a larger strategic partner, already equipped with capital and operating resources, to take the platform forward.

Both paths were credible.

The company’s partners had seen enough technical progress to support further expansion. They wanted the team to move ahead with a capital plan and increase production capacity.

But building that capacity would have required fresh capital, creating the potential for significant dilution.

However, the decision was not purely financial.

The capabilities required to develop a material at bench, research, and demonstration scale are very different from those required to own and operate commercial manufacturing assets.

A startup may be confident that it can build a plant. Operating one safely, reliably, and in an environmentally sound way is a different challenge. It requires different people, systems, and daily disciplines.

The team understood that bringing those capabilities in-house would be difficult and costly. The skills that had made the company successful in technology development would not automatically make it an effective manufacturing operator.

That distinction became central to the strategic decision, and rather than viewing an acquisition as the automatic end goal, the company treated it as a threshold calculation.

A final lesson from this journey is that the decision reflected the same discipline that shaped the company’s path: understanding what it was uniquely capable of building, recognizing where external partners could create greater leverage, and choosing the option that best preserved the value of the technology and the people behind it.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape effectively.

For twenty years, the dominant technology companies trained investors to think of scale as a matter of software distribution. A product could be built once and sold globally. Marginal cost could fall toward zero. Cloud computing could turn fixed capital into variable expense. App stores, APIs, digital advertising, and payment rails allowed small teams to act on large markets without owning much physical infrastructure.

That era is not over. But its institutional assumptions are now misleading. Artificial intelligence, advanced manufacturing, robotics, grid-scale batteries, energy storage, data centers, semiconductor supply chains, synthetic fuels, drones, heat pumps, and industrial automation do not scale like consumer software. They require land, permits, power, cooling, transformers, chips, metals, factories, spare parts, operating crews, insurance, and long-term contracts. Each additional deployment often requires additional capital.

Subscribe now

Venture capital was designed for the discovery phase of this world. It can fund a team before the product works. It can absorb technical risk. It can price extreme upside. It can create a market around a story that might become true.

But once a physical technology starts working, the problem changes. The question is no longer how to finance the first product. It is how to finance the next thousand units.

This is the hard middle: too late for pure invention capital, often too early for ordinary bank debt. The system works, but the asset class is not yet familiar. A company that keeps raising equity to finance repeatable assets may still have excellent technology. But it has not yet built a deployment engine.

This is the reason the next industrial titans will not look like ordinary startups. They will be engineering companies with capital engines. Their advantage will not be only that they know how to build a machine. It will be that they know how to turn the machine into a repeatable, underwritable unit of industrial capacity. By integrating a multi-way capital stack into the production system in concert with variables such as engineering, procurement, regulatory strategy, and insurance.

From here onward, only our Premium Members can keep reading.
The rest of this piece explains why this shift is already happening, which new funding models are beginning to emerge across geographies, sectors, and real companies.

Read more

Welcome to Edition No. 113 of Deep Tech Briefing.

The Scenarioist’s weekly intelligence layer on what you need to know, understand, and track to win in deep tech: company milestones, market shifts, and macro forces reshaping outcomes, competitive position, capital allocation, and critical decisions.

In this edition, the Big Idea is Hermeus. Its Mach 1.21 flight is the headline, but the deeper question is whether high-speed aerospace can become a repeatable industrial market — one where capital, procurement, testing infrastructure, supplier learning, and mission demand start reinforcing each other.

Beyond it, more than 30 company milestones were analyzed across space, AI infrastructure, quantum, nuclear, manufacturing, materials, food-tech, robotics, autonomy, and risk infrastructure.

The market-making layer analysis this week moves from Europe’s sovereign AI push and allied AI-security coordination to subsea cable protection, advanced nuclear fuel access, and New Zealand’s maritime defense expansion.

And finally, the edition closes with a curated selection of 10 key data points to know — from 64 ASICs per AI rack to ~20 metric tons of strategic nuclear material.

Deep Tech Briefing is available for Premium Members

By joining The Scenarioist Premium, you also unlock access to Deep Tech playbook, exclusive lessons, exit analyses, scalability frameworks, and hard-earned insights from more than 100 founders and investors.

Unlock Premium Access

Speed Becomes an Asset Class

Hermeus’ Mach 1.21 flight forces a harder question: can high-speed aerospace become a repeatable market for capital, test data, procurement, and industrial capacity?

Read more

Subscribe now

This week, I sat down with Alan Yu, Senior Associate at Space Capital, to unpack how satellites are becoming one of the invisible operating layers of the broader space economy, and what founders need to know about building, validating, and funding successful companies in the sector.

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

The Space Economy Is Becoming an Operating Layer of Modern Life

Framing the satellite economy in 2026 requires moving beyond the image that still dominates much of the public imagination.

For many people, space still means rockets, astronauts, missions, or assets orbiting far above Earth. That view is not wrong, but it is increasingly incomplete.

A more useful way to understand the space economy today is to see it as an operating layer of modern life.

Satellites are no longer just instruments of exploration. They have become part of the infrastructure that supports transportation, communications, financial systems, mapping, climate intelligence, defense, logistics, and many other activities that now depend on space-enabled capabilities.

Most people do not experience this directly as “space.” They experience it as a phone that knows where it is, a plane that remains connected in the air, a ride-hailing app that works, a bank transaction that relies on precise timing, a map that updates, or a system that can observe conditions on Earth from above.

That is what makes the “satellite economy” strategically important.

Its value is not limited to what happens in orbit. Its value comes from the way orbital infrastructure becomes embedded into terrestrial markets.

The more these systems work, the more invisible they become. And the more invisible they become, the more dependent the economy becomes on them.

Satellites as invisible infrastructure

The satellite economy can be understood through three major capabilities: satellite communications, GPS and precision navigation and timing, and geospatial intelligence.

Each of these categories has a different history, a different technical architecture, and a different set of commercial applications. But together they reveal the same underlying pattern: space-based infrastructure has become a utility.

That utility-like nature matters.

Once society has access to these systems, it does not want less of them. It wants more coverage, more resilience, more precision, more bandwidth, and more intelligence.

No one expects GPS to disappear from daily life. No one wants connectivity to become less available. No institution wants timing systems to become less reliable.

The more digital and autonomous the economy becomes, the more important these satellite-enabled systems become.

GPS is the clearest example.

For consumers, GPS often appears as a convenience. It helps a phone identify a location, supports navigation, enables ride-sharing, and makes mapping applications feel effortless. But that consumer-facing experience represents only a small part of the underlying importance.

Precision navigation and timing reach much deeper into the economy. They support transportation networks, autonomous systems, logistics, and financial infrastructure.

Satellite communications have followed a similar path.

For a long time, satellite connectivity was associated with specialized use cases: aircraft, ships, remote locations, or government and defense environments. It was valuable, but not necessarily something most people thought of as a broad consumer or enterprise infrastructure layer.

That perception has changed.

With the rise of large satellite constellations and more ambitious connectivity models, satellite communications are becoming part of a much broader conversation about global access, redundancy, mobility, and resilience.

The expectation is no longer that connectivity should exist only where terrestrial infrastructure is convenient. Increasingly, the expectation is that connectivity should extend almost everywhere.

Geospatial intelligence adds another dimension.

Earth observation has long been one of the most important uses of satellites, but the strategic value is not in imagery alone. The value comes from the ability to understand what is happening on Earth and translate that understanding into decisions.

That distinction is essential.

A satellite image by itself is not always useful to the final customer. A raw signal does not automatically create economic value.

Between the capture of data and the moment when a user can act on it, there is a long chain of processing, interpretation, contextualization, and distribution.

That is why geospatial intelligence sits at the intersection of infrastructure and application. It begins with orbital assets, but its value is realized only when the information becomes usable for someone on Earth.

This is one of the central tensions in building space tech startups. The infrastructure is technically complex, but the customer often wants simplicity.

The system may involve satellites, sensors, signal processing, coordinate systems, imagery pipelines, AI models, and distribution platforms. But the buyer cares about the decision the system enables.

That is where much of the entrepreneurial opportunity now begins to appear.

From space assets to a broader space economy

The deeper shift is that the space economy can no longer be defined only by the assets that physically operate in space.

Those assets remain foundational. Rockets, satellites, propulsion systems, ground stations, sensors, and orbital logistics are still essential.

Without them, there is no infrastructure layer to build on. But the market is larger than the hardware in orbit.

The space economy now includes the companies that build and operate satellites, the firms that move and manage assets in space, the platforms that process data, the software systems that distribute intelligence, and the businesses that turn space-derived capabilities into products for Earth-based customers.

That broader view is important because it changes who can build in the sector.

Historically, space was accessible to a narrow group of actors.

It required enormous capital, specialized engineering talent, government relationships, and tolerance for long development timelines.

Those realities have not disappeared, especially for companies building core infrastructure. But the ecosystem has expanded, with direct and indirect implications for founders.

First of all, a founder does not always need to build the entire satellite system.

A company may focus on a specific sensor, a specific mission, a specific data layer, a specific distribution problem, or a specific customer workflow.

As manufacturing improves, launch costs fall, satellite buses become more available, and the surrounding ecosystem matures, the opportunity space becomes more modular.

That does not make space easy. It does make it more strategically open.

Venture design should now begin with a sharp question:

“Which part of the stack is essential for us to own, and which part can be accessed through the ecosystem?”

That question is central to company building in the current satellite market. It forces founders to distinguish between technical ambition and strategic necessity.

There is a difference between building something because it is impressive and building it because the company cannot create durable value without owning it.

The more capital-intensive the market, the more important it becomes to understand where true differentiation lives.

The satellite economy is therefore not just expanding in size. It is changing in structure. One useful frameworks that emerged from the conversation is the three-layer view of the satellite economy:

1. At the base is foundational infrastructure: launch, satellite manufacturing, propulsion, orbital logistics, ground systems, communications, and the physical systems that make space operations possible.

2. In the middle, there is distribution: the processing, filtering, interpretation, packaging, and delivery of space-derived data and capabilities into forms that customers can actually use.

3. At the application layer, there are businesses that build products on top of those capabilities for agriculture, insurance, defense, mobility, logistics, autonomous systems, financial services, urban planning, and many other markets.

This layered view matters because each layer behaves differently. At a high level:

• Infrastructure companies often require deeper technical teams, more capital, longer development cycles, and stronger execution against physical systems.

• Distribution and application companies may be more software-driven, but they still depend on the maturity, availability, reliability, and cost structure of the underlying space infrastructure.

The pitfall is assuming that infrastructure availability automatically means market readiness.

A satellite may be able to capture the data. A provider may be able to make that data available. A founder may be able to build a product concept around it.

But the company still has to prove that the cost structure works, that the customer understands the value, that the insight can be delivered frictionlessly, and that the willingness to pay is real.

That is why satellites should not be viewed only as a technology category. They should be viewed as the foundation of a developing business architecture.

The current moment is compelling because several things are happening at once.

• Satellites are becoming more capable.

• Launch and manufacturing are becoming more efficient.

• Connectivity is expanding.

• Sensors are improving.

• Compute is moving closer to orbit.

• The number of companies contributing to the ecosystem is growing.

• And customers on Earth are becoming more aware that space-derived intelligence can support real operational decisions.

But the market has not reached full maturity.

The infrastructure is advancing faster than many applications can absorb. The capabilities are becoming stronger, but the translation into customer value remains uneven. In other words, the potential is broad, but it is not automatic.

That is the right starting point for understanding the satellite economy in 2026.

Space is no longer a separate frontier sitting outside the rest of the economy. It is becoming a layer beneath it. The companies that matter most will not simply be the ones that put assets in orbit. They will be the ones that understand how orbital infrastructure becomes usable, valuable, and investable on Earth.

Satellites Are Becoming Smarter

As discussed, satellites are becoming more capable, more coordinated, and more useful as operational assets.

They are no longer just passive machines placed in orbit to capture images, transmit signals, or provide connectivity.

Increasingly, they are becoming intelligent infrastructure: machines that can compute, filter, coordinate, move, and eventually operate as part of larger systems.

That distinction matters because it changes the strategic role of satellites in the space economy.

A satellite used to be thought of as a highly specialized node. It had a mission, a payload, a communications function, and an orbital position. It was expensive to build, expensive to launch, difficult to move, and often limited in flexibility once deployed.

That model is giving way to something more dynamic.

The next generation of satellites is not simply cheaper. It is becoming more operationally relevant.

The cost of building and launching systems with comparable capability has fallen, but at the same time the capability of those systems has increased dramatically.

Better manufacturing, stronger sensors, more powerful onboard compute, and more flexible propulsion are beginning to reshape what satellites can do and how companies can build around them.

From static assets to autonomous machines

The most important change in satellite infrastructure is the move from static orbital assets toward more autonomous systems.

This does not mean every satellite suddenly becomes fully independent or intelligent in the way people sometimes imagine when talking about AI or robotics.

The point is more practical.

Satellites are gaining the ability to do more of the work closer to where the data is generated.

Onboard compute is a simple but powerful example. If a satellite captures imagery, not every image has the same value. Some data may be redundant, cloudy, low quality, or irrelevant to the customer’s need.

In older architectures, much of that data would need to be transmitted down and processed later, creating cost, latency, and inefficiency.

As satellites gain more compute capability, they can begin to filter and prioritize information in orbit. They can help decide what is worth sending down and what is not. That changes the economics of data movement and the usefulness of the system.

It also changes the meaning of satellite infrastructure itself.

The satellite becomes less like a remote instrument and more like part of a a computing-enabled asset inside a broader space infrastructure layer.

It is still a physical asset in orbit, but it increasingly participates in the intelligence layer of the system. It does not merely collect information. It helps make information more usable.

That matters as the volume of data increases.

More satellites mean more coverage, more sensors, and more raw information. But more information does not automatically create more value. Without filtering, processing, and distribution, the system can produce complexity faster than it produces insight.

The smarter the satellite becomes, the more it can help reduce that friction.

This will matter beyond Earth observation. It will matter for communications, defense, orbital coordination, robotics, lunar operations, and any environment where space assets need to work together in real time.

The broader direction is clear: satellites will need to coordinate with other satellites, ground stations, logistics providers, defense networks, data platforms, and end-user applications.

The more congested and strategically important the orbital environment becomes, the more valuable that coordination becomes.

This is why the satellite of the future should not be understood only as hardware. It should be understood as a node in an increasingly complex operating system.

The New Satellite Stack Changes What Founders Can Build

Historically, building a space company often meant confronting the full burden of space infrastructure.

A founder who wanted a specific type of data or sensing capability had limited options. If the required satellite data did not exist, or if existing providers could not meet the need, the company faced an enormous problem.

Building the full system meant designing spacecraft, sourcing components, managing payloads, dealing with launch, handling operations, and raising significant capital before proving much about customer demand.

That is no longer always the case. Here are a few reasons, drawn from our conversation:

• The ecosystem has become more modular.

• Manufacturing capacity is improving.

• Sensors and power systems are becoming more accessible in certain categories.

• Launch remains difficult, but lower launch and manufacturing costs have changed what early-stage teams can realistically attempt.

• Specialized companies can now take on parts of the stack that every founder previously might have been forced to confront directly.

That changes the early-stage roadmap.

A company may have a differentiated sensing concept, a specific customer insight, or a mission-driven application without needing to build every piece of the space system from scratch.

It can focus on the part of the architecture that matters most to its strategy and rely on partners or suppliers for the rest.

This is a profound venture-building shift.

It also means founders need to be much more precise about what they choose to own.

Owning more infrastructure can create control, defensibility, and performance advantages. But it also increases cost, complexity, hiring burden, development time, and capital requirements.

In a market where timing and capital efficiency matter, unnecessary ownership can become a liability. The best founders understand this tradeoff early.

Some companies should absolutely be infrastructure companies.

They need deep engineering teams, flight heritage, physical systems expertise, and the ability to build hard technology that works in space. These companies may require more capital because the problem itself demands it.

But many other companies in the space economy do not need to own the whole stack. They may be building intelligence products, data platforms, customer workflows, analytics layers, or application-specific systems that depend on space infrastructure without being pure infrastructure companies themselves.

For them, the advantage may come from customer understanding, distribution, workflow integration, or the ability to turn complex signals into actionable insight.

That difference matters to investors.

Capital intensity is not automatically bad in Deep Tech. Some problems require capital because the technical challenge is real and the reward is large enough to justify the effort. But capital intensity must be matched by the right kind of value creation.

In the end, the changing satellite ecosystem gives founders more room to build. But it also forces them to decide early what kind of company they are actually building.

• Are they building a core infrastructure company?

• Are they building a distribution layer?

• Are they building an application business that depends on space-derived intelligence?

• Are they creating a new capability that requires owning hardware, or are they using existing infrastructure to solve a customer problem more effectively?

Those questions shape the team, the fundraising strategy, the roadmap, the customer discovery process, and the investor narrative.

Founders Should Build Only What Truly Matters

One of the most important disciplines in Deep Tech company building is knowing what not to build. That sounds simple, but in technically ambitious markets it is one of the hardest judgments to make.

The best founders are not the ones who try to own everything. They are the ones who understand which part of the stack defines their advantage.

Core space infrastructure companies

For companies building core space infrastructure, the technology simply has to work.

There is no way around that. A satellite company, propulsion company, orbital logistics company, sensing company, or hardware-driven space startup cannot compensate for weak technical execution with storytelling alone.

The team has to understand the engineering reality.

The company has to attract people who know how to build hard systems, ideally with experience that reflects the practical difficulty of getting technology to operate in space.

Therefore, a founder does not necessarily need to be the most technical person in the company, but the founder does need to understand the technology deeply enough to lead the organization, recruit the right people, ask the right questions, and avoid being disconnected from the technical truth of the business.

That is different from simply having a technical background.

• It requires enough fluency to make strategic decisions around tradeoffs, timelines, architecture, risk, and customer fit.

• It requires the ability to work closely with engineers without reducing the company to an engineering project.

This is where many Deep Tech founders encounter a subtle trap. Because the technology is difficult, it can become the center of gravity of the entire company.

Every milestone becomes a technical milestone. Every discussion returns to the next engineering challenge. The company keeps moving toward better technology, but not necessarily toward a better business.

That is not enough.

In venture-backed Deep Tech, the goal is to build a system that matches what the customer needs at the moment when the customer is ready to adopt it.

If this direction is not managed carefully, the path can consume years, force unnecessary fundraising, dilute the company, and make the team believe it is making progress while the market remains unproven.

Distribution and applications

For founders building distribution systems or applications on top of satellite infrastructure, the risk is different.

These companies may not need to build satellites, launch systems, or orbital vehicles themselves.

They may be using existing data, connectivity, geospatial intelligence, or distribution systems to create products for specific markets.

In principle, that should make them less capital intensive and more flexible. But the risk is assuming that the underlying inputs are ready simply because they exist.

A founder may see an opportunity in agriculture, climate, insurance, logistics, defense, autonomous systems, or infrastructure monitoring.

The use case may be compelling. The customer problem may be real. The value of space-derived intelligence may be easy to imagine.

But the company still has to prove that the inputs can support the business.

• Is the data available at the right frequency?

• Is it accurate enough?

• Is it affordable enough?

• Can it be processed reliably?

• Can it be delivered into the customer’s workflow without too much friction?

These questions determine whether an application business can actually exist.

In geospatial intelligence, this issue appears repeatedly. Satellite imagery can be powerful, but the cost of acquiring, processing, and translating that imagery into something useful can still be significant.

In some cases, price is a constraint. In others, the deeper constraint is that customers do not yet understand how to use the intelligence in a way that changes behavior.

That distinction matters.

For application founders, the work is therefore not just technical development. It is market development.

They must prove that the infrastructure they depend on is mature enough to support their use case, and that the customer is ready enough to turn that capability into revenue.

Cost structure shapes the business model

The cost of data, infrastructure, talent, manufacturing, deployment, customer support, integration, and technical iteration determines how much flexibility a company has.

It affects pricing. It affects margins. It affects the sales motion. It affects whether the company can experiment with different customer segments or whether every customer has to be large enough to justify the effort.

A business model is not a spreadsheet exercise performed after the product has been defined. It is shaped by the architecture of what the company chooses to own, what it chooses to avoid, and how directly its costs map to customer value.

In space-enabled markets, this is particularly important because the cost of inputs can be substantial.

Satellite imagery, data processing, infrastructure access, customer integration, and domain-specific interpretation can all shape the economics of the company.

Simple, but not obvious:

• If those costs are too high, the company loses flexibility. It may need to charge more, sell only to larger customers, or narrow its market to the few buyers who can justify the price.

• If the costs are lower, the company has more strategic room. It can experiment with pricing. It can serve more customer types. It can adjust the sales motion. It can protect margins while learning. It can avoid forcing the market into a pricing structure it is not ready to accept.

That is why cost drives the business model more than founders sometimes want to admit. The product may define what the company does, but the cost structure defines what the company can afford to become.

One of the most effective ways to approach this problem is to work backward from the customer.

Not backward from the technology. Not backward from the product vision. Not backward from the total addressable market slide. Backward from what the customer values enough to pay for.

That requires a founder to ask uncomfortable questions early.

• How much would this customer actually pay?

• What would make them pay more?

• What would make them stop paying?

• What part of the product creates the strongest pull?

• What existing tool or process would this replace?

Capital Efficiency Begins With the Right Founder Mindset

Some space companies genuinely need significant capital.

A company building orbital logistics, propulsion, energy transfer, satellite infrastructure, or another deeply technical system may require specialized talent, expensive testing, and enough runway to reach a meaningful technical milestone.

In these cases, every dollar should be tied especially clearly to risk reduction and company value.

Capital efficiency is not simply about spending less. It is about knowing what the next dollar is supposed to prove. It is about separating essential progress from motion, and understanding which risks deserve capital now versus which can be tested through sharper prioritization, better sequencing, or a more focused roadmap.

Founders may be tempted to raise more than they need.

That can feel rational. More capital seems to mean more safety, more hiring power, more technical ambition, and more room to maneuver. But over-raising can create its own risk.

When a company raises too much too early, it can start behaving as if the next stage has already been earned. It hires ahead of clarity. It expands the roadmap before the company knows which milestones really matter. It becomes easier to hide weak signal behind a larger budget.

Early-stage companies do not need to prove everything.

They need to prove the right thing.

The founder’s task is to understand what the next financing round, customer commitment, technical milestone, or strategic inflection point will depend on.

The amount of capital raised should reflect that path.

It should be enough to reach the next proof point with credibility, but not so much that the company loses urgency, accumulates unnecessary complexity, or creates expectations it cannot support.

Capital is not the strategy. Capital is the fuel for a strategy that has to be clear before the money arrives.

Early teams should stay flexible

Capital efficiency also shows up in how a founder builds the team.

Early-stage companies often feel pressure to look complete. They want defined executive functions because that structure resembles a mature company. But early startups are not mature companies. They are learning systems.

At the earliest stages, rigid roles can become premature.

The company may not yet know what kind of operations it truly needs, what kind of sales motion will work, how technical development will evolve, or which customer segment will define the first real market.

That is part of what makes venture building in Deep Tech difficult.

The problems are often undefined. The markets may still be forming. The technical architecture may change as the company discovers what is feasible, valuable, and affordable.

Because of that, every company needs to identify its own catalyst.

For some companies, the key metric may be revenue.

For others, revenue may matter less than a specific customer commitment, a technical demonstration, a signed partnership, a successful test, a reduction in cost, or evidence that a particular buyer is engaging seriously.

The important thing is to know what evidence will change the probability of success.

That mindset should show up in investor updates as well.

A company can share product developments, customer conversations, technical progress, hiring updates, partnerships, experiments, and operational work.

All of that can be useful. But the best updates reveal focus. They show what the company is learning, where the team is concentrating effort, and how the business is extracting more value from the resources it already has.

That input-output discipline is a powerful signal.

It shows that experimentation is not random. It is tied to a strategic hypothesis about what will make the company work.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape effectively.

The Scenarionist Premium is designed to make you a better Deep Tech professional. Premium members gain exclusive access to unique insights, analysis, and masterclasses with the wisdom of the world’s leading Deep Tech thought leaders. Invest in yourself, and upgrade today!

Subscribe now

Deep Tech companies rarely fail for one reason alone.

Sometimes the science does not work. Sometimes the market is not ready. Sometimes the cost curve is worse than expected. Sometimes the first application is wrong. Sometimes the technology is real, but the customer cannot adopt it. Sometimes the company raises money around a roadmap that reality refuses to follow.

But across the conversations I have had with founders, investors, operators, and industry experts in Deep Tech, one pattern keeps coming back.

Deep Tech companies also fail when the wrong people own the wrong problems at the wrong time.

At first, this may sound like an organizational issue.

I don’t think it is.

It is a company-design issue.

A Deep Tech startup is not simply a lab with a pitch deck attached to it. And it is not a software company with a longer technical roadmap.

It is a fragile system where scientific risk, market risk, operational risk, financing risk, and governance risk begin to interact very early.

Subscribe now

At the beginning, the company is still mostly the team.

There is no mature organization to absorb ambiguity. There are no established processes to compensate for unclear ownership. There is no predictable commercial engine, no industrial machine, no stable deployment motion, and often no meaningful revenue.

What exists is a small group of people trying to turn a technical insight into a company that deserves to exist.

A young Deep Tech company can have a CEO, a CTO, advisors, investors, and a few impressive profiles, while still leaving the most important risks under-owned.

It can look strong from the outside and still be fragile inside.

An interesting way to frame this issue is to ask:

“Which critical jobs have to be owned for the plan to succeed?”

That is the question I want to explore in this piece.

Not the generic statement that “team matters.” Everyone says that.

The more useful question is how the team actually functions when it has to translate science into product, product into adoption, adoption into delivery, delivery into economics, and economics into a fundable company.

That is where Deep Tech teams win or break.

The Framework We’ll Follow Today

There are many ways to discuss teams in Deep Tech. We could talk about founder-market fit, leadership style, hiring, culture, governance, or incentives.

All of those matter.

So in this piece, I want to look at the team question through three practical angles, drawing on a curated narrative built from the conversations collected so far.

The goal is to build a sharper language for asking better questions: who owns which risk, where does misalignment begin, and how can the team fine-tune itself before small gaps become expensive.

Subscribe now

1. What Winning Teams Get Right

Before an early-stage Deep Tech company has a real organization, it usually has a small number of people carrying a large number of tasks.

Costs begin on a clear date. Revenue, instead, arrives on an uncertain one.

That is why roles must be defined.

However, team dysfunctions tend to show up in a few recurring ways. Across our work, a few patterns keep appearing:

Read more

Picture a city bridge scheduled for replacement.

To the public budget, it is a cost. To the contractor, it is logistics. To the financial system, it is mostly invisible.

But inside that bridge are steel, rebar, asphalt, copper, lighting, guardrails, electrical components, and signage. Some of it is waste. Some of it is scrap. Some of it could be reused, recertified, insured, bought, and sold.

That is where this week’s Deep Tech Briefing starts: with an interesting provocation from a Dutch company.

The full The Big Idea looks at where the real value sits: not in generic reuse, but in the trust layer that lets buyers, insurers, lenders, and procurement teams make real decisions.

Subscribe now

Going beyond that, this week’s milestones point to the same underlying question: when does a hard asset, or a hard technology, become legible enough for the market to act?

A few signals stand out: infrastructure financing around compute and autonomy, advanced-energy partnerships focused on manufacturing and construction, and industrial AI applications moving closer to physics-based and regulated environments.

The same issue also looks beyond company milestones, where defense procurement, AI classification, critical-minerals governance, and PFAS policy are beginning to shape the market before the outcomes are obvious.

Enjoy the read.

✨ Deep Tech Briefing is the weekly intelligence layer that reads the global deep tech ecosystem through the signals that matter: milestones, contracts, deployments, policy shifts, financing structures, industrial bottlenecks, and failures.

If you build, back, or study Deep Tech, it is designed to sharpen your lens — and maybe help you avoid learning a few lessons the expensive way.

Subscribe now

There Is No Reindustrialization Without Materials Accounting

Recoverable materials become financeable only when the built environment has a ledger.

Before factories can become more productive, the physical inputs around them must become visible, priced, and reusable.

Every industrial strategy begins with a promise to build. Factories, grids, shipyards, data centers, housing, battery plants, and defense production all require physical inputs. Steel must arrive. Copper must arrive. Concrete, aluminum, asphalt, glass, timber, fixtures, and machinery must arrive.

The odd part is that much of this material is already here. It is sitting in buildings, roads, bridges, schools, ports, warehouses, municipal assets, and utility systems. It has already been mined, refined, transported, financed, installed, insured, and depreciated. Yet the financial system usually notices it only when somebody pays a contractor to tear it out.

This week, a Netherlands-based organization working on the financial layer for circular construction put forward a useful provocation:

Read more

Subscribe now

This week, I sat down with Haig Farris, Co-Founder of D-Wave, to unpack what it takes to build a successful Deep Tech startup in quantum computing when the technology is ahead of the market, the product path is uncertain, and the commercial outcome may take decades to fully emerge.

Founded in 1999, D-Wave reached the public markets in 2022. Today, we’ll explore how the company began as a startup at the frontier of quantum technology and took its first steps toward commercialization through technical progress, strategic pivots, patient capital, and excellent teamwork.

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

Starting a Quantum Company Before the Market Was Ready

Framing the story of D-Wave requires starting from a point that differs from the usual founder narrative around frontier science.

It did not begin with a market that was already waiting for a quantum computer. And it certainly did not begin with a clean, obvious path from scientific insight to commercial product.

It began with people.

That distinction matters because in the earliest stages of Deep Tech, especially in a field as new and difficult as quantum computing was in 1999, investors often are not underwriting a market-ready product. They are underwriting exceptional people who may be able to make the technology real.

That was the case in this story.

The company began with people before it began with a product

As we discuss every week, in Deep Tech, the ability to translate complexity is not a soft skill. It is part of the company’s commercial infrastructure.

If a founder cannot explain the core idea to investors, early employees, strategic partners, and eventually customers, the science remains trapped inside the lab.

From the beginning, D-Wave was not built only around deep physics. It was built around people who could make that physics understandable, credible, and commercially meaningful.

What stood out was the combination of deep physics and commercial vision: the ability to see not only a scientific breakthrough, or the possibility of one, but why it might matter outside academia.

At that stage, there was no mature quantum computing market waiting for a product.

What existed first was the ability to make a complex idea compelling enough for others to believe it could one day become economically important.

That is the first lesson here.

The scientist-entrepreneur has to be capable of translating the impossible into stages of progress. And the company has to keep accumulating value before the final commercial outcome is clear.

The longer the commercialization path, the more the team must be unusually smart, unusually driven, and unusually capable of bringing others into a difficult vision.

The founders must show that they can keep learning, keep attracting talent, and keep making progress even when the market is not yet ready.

In that sense, D-Wave did not start because the market was ready. It started because the people were unusual enough to make the market worth waiting for.

IP became an asset before the product became obvious

Another key aspect of this journey is that the intellectual property D-Wave accumulated in its early years became one of the company’s most important assets.

In a science-intensive company, IP can become part of the financing story before the product is commercially mature. That becomes especially important when the time to market is measured in years.

For D-Wave, the patent portfolio helped investors believe that even if the company did not reach the commercial outcome everyone hoped for, there was still real underlying value.

The company had built something defensible. It had accumulated ownership around a field that could become important. It had a position that might retain value even in a downside case.

That was especially important because the early product thesis was still uncertain.

Quantum computing was not a market with established demand. There were no standard customers buying quantum machines in the ordinary way. The company was still trying to make progress on the science and engineering, while also convincing investors that the value of the company was increasing.

The patent portfolio helped make that argument.

That is not the same as saying patents alone made the company investable. They created a layer of value that made the long journey more fundable.

This is an important distinction for Deep Tech founders.

When the Original Thesis Does Not Work as Planned

One of the most useful parts of the D-Wave story is that the original pitch did not work. It is central to understanding how long-horizon Deep Tech companies actually survive.

At the beginning, there was a belief that a certain kind of technical demonstration would be enough to create strategic pull.

If the company could demonstrate that it could link two qubits together and show how they functioned, the large technology companies would understand the importance of what had been built.

The expectation was that the field’s major players would beat a path to the company’s door.

That did not happen.

The team did the work. They linked the qubits. They published papers. They gave demonstrations. The milestone was real. But the market reaction was not what the company had expected.

No large incumbent arrived with the urgency the founders and investors had imagined. No strategic buyer treated the demonstration as the inevitable beginning of a large commercial opportunity.

So the company had to raise more money and take the next step.

That sentence captures a great deal of what frontier company-building feels like.

A technical milestone can be genuine and still not be enough. A scientific proof point can be important and still fail to unlock the next commercial chapter. The market can be impressed without being ready. Large companies can observe the future without feeling the need to own it immediately.

In a conventional startup, that mismatch might be fatal.

If the company is built around a short-term commercial assumption and the customer does not respond, the business may not have enough reason to continue.

But in a frontier technology company, the first commercialization thesis can be wrong while the underlying company still becomes more valuable.

That is what happened here. The company kept going because it kept becoming better.

The Strategic Pivot from Gate Model to Annealing

There is a moment in some Deep Tech companies when the original technical path does not disappear because it is wrong, but because it is too far away.

That distinction is important.

The early D-Wave journey began with an ambition around quantum computing that was enormous in scope. The company was working on a gate-model path, a broader approach to quantum computing than the annealing route it would later choose.

But after roughly five years, the practical reality became impossible to ignore. The gate-model approach was going to take too long.

The problem was that the technical obstacles made that path look too long for the company to keep pursuing as its main route. There were major issues around noise.

There were major issues around linking qubits in large numbers. The path might have been scientifically meaningful, but from a company-building perspective, it was too distant.

That was the strategic significance of D-Wave’s pivot to annealing.

Focus came from accepting what would take too long

Broad visions can become dangerous when the technical execution horizon is too long. A startup does not have infinite time. It does not have infinite money.

The annealing approach was described as a faster, more practical way of getting to a quantum computer that worked in a particular domain: optimization.

It was narrower. But it was reachable.

In brief, annealing offered a different logic: instead of continuing to pursue the broader gate-model route as the company’s main path, the company could concentrate on a class of problems where quantum effects could be applied more directly.

Optimization became the center of gravity.

That meant the company could focus its technology development and financing story around a more specific kind of value proposition.

That is one of the most difficult strategic moves in Deep Tech: knowing when to stop pursuing the original vision and start pursuing the version that can become a company, especially when there is no existing path to follow.

The company changed direction overnight

The pivot was decisive.

Once the company switched to the annealing path, it did not simply add annealing as another research direction. It changed the company.

The organization moved away from the earlier path because the team believed it already knew how to make the chips required for the annealing approach.

The decision also reshaped the financing story. After the pivot, the company began more serious attempts to raise capital and eventually found the right investors.

That financing did not happen in a vacuum. It came after the company had chosen a more focused technical path and could tell a clearer story about what it was trying to build.

In that sense, annealing was not merely a technical pivot. It also reshaped the business architecture of the company. It defined what kind of problems the company would focus on. It defined what kind of investor story could be told. And it allowed the company to move toward a more focused path for engineering and commercialization.

From Scientific Breakthrough to Commercialization

The technical journey of D-Wave was not only about making a quantum machine work for a specific class of problems.

It was about turning a deeply complex physical phenomenon into a system that could keep improving, solve increasingly difficult problems, and eventually be used by people who were not quantum physicists.

That distinction matters because in Deep Tech, the core breakthrough is rarely the whole product. It is the beginning of the product.

In the case of D-Wave, the technical obstacles can be understood through a few central problems:

• Reducing the number of wires needed to link qubits together.

• Reducing noise so the quantum state could be maintained.

• Linking more qubits together so the machine could address large optimization problems.

To link qubits together, the company had to send wires down into the system. Reducing the number of wires needed to create those connections was a major step.

In a normal technology company, a wiring problem might sound secondary. In this context, it was central because the ability to connect qubits was part of the path toward making the machine capable of doing something powerful.

The second challenge was noise.

A quantum state is delicate. If the electronics create too much disturbance between the chips, the quantum state can be disrupted. If the quantum state is disturbed, the calculation is not accurate. That means noise is not only a performance issue. It directly affects whether the computation can be trusted. The machine has to protect the state well enough for the computation to matter.

The third challenge was scale and connectivity.

The company had to learn how to link more qubits together because that is where the computational power begins to matter.

Low noise, many qubits, and deep connectivity between them were the ingredients that could allow the system to process large optimization problems.

This is where the technical progress became inseparable from the business story.

The company was not simply making a better chip for the sake of making a better chip. It was improving the system so that the machine could solve more difficult problems.

That is the way technical progress becomes commercial progress in frontier hardware.

Another essential part of the system was software.

For the technology to become useful, users with optimization problems had to be able to work with the machine without first becoming quantum physicists.

That meant building software that could translate a real optimization problem into the quantum requirements of the system.

Building a Company That Can Outlast Its First Founders

One of the most interesting lessons from the D-Wave story is that a long-horizon Deep Tech company cannot depend forever on the same kind of talent that made it possible in the first place.

The earliest phase requires people who are willing to live at the frontier.

These are the people who are excited by uncertainty, who want to work on problems before the path is clear, and who can operate in a world where the difference between a technical enabler and a company strategy is still being worked out.

In D-Wave’s case, that kind of talent was essential at the start. Without it, there would have been no company to build.

But as the company moves forward, the work changes.

At some point, the problem is no longer only whether the science can be imagined. It becomes whether the technology can be engineered, financed, sold, supported, and managed over many years.

The company has to move from invention to execution. And that transition can require a different kind of organization, with different kinds of talent.

Long time-to-market companies need belief, adaptation, and luck

Another advice for founders building in a long time-to-market sector is both simple and difficult: keep at it, but only if the belief is grounded in real economic value.

That distinction is essential.

It is not enough to believe that the science is beautiful. It is not enough to believe that the technology is intellectually important. A company can only keep raising money, attracting talent, and surviving hard cycles if the founders and investors believe that the work has long-term economic significance.

The value has to be more than scientific.

In D-Wave’s case, the belief was that quantum computing could eventually matter economically because of the kinds of problems it could address.

Even when the original commercialization assumptions did not play out, the company could continue because investors believed the underlying asset was becoming more valuable.

The technology improved. The IP improved. The team improved. The company kept moving.

That is the foundation of persistence in Deep Tech.

Persistence is not simply refusing to quit. It is continuing to make the company more valuable while the world catches up to the technology.

But persistence alone is not enough.

The company also has to adjust to the times. It has to adjust to the people it has, the financing environment it faces, and the market it is trying to reach. The strategy cannot be frozen at the moment of founding.

The company has to keep understanding the market well enough to know how to pitch itself, what kind of people to recruit, and how to move the business forward.

That is where the board becomes important.

In a long-horizon Deep Tech company, the board is not just there for governance. It has to help the company survive transitions.

It has to help find the right leadership. It has to help the company recognize when the talent needed for the next phase is different from the talent that defined the first phase. It has to help assemble the team that can carry the company forward.

In the end, a long time-to-market company is not built by a single act of genius. It is built through repeated acts of continuation.

The company has to survive the period when the market is not ready, when the original pitch does not work, when the technical path changes, when founders move on, when investors need reassurance, and when the product is still becoming understandable to customers.

In that kind of company, belief is necessary, but belief has to be renewed through progress.

The company has to become more valuable every year. Not always through revenue at first, and not always through commercial validation, but through capability, defensibility, talent, and strategic relevance.

D-Wave’s story is therefore not only a quantum computing story. It is a story about what it takes to build a company in a field where the market may take decades to fully form.

The technology has to advance. The people have to evolve. And the organization has to become strong enough to carry the company forward, even if the original thesis has evolved.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas.

Rumors is a pattern-recognition layer for investors, founders, and operators who care about where the frontier is heading before it becomes consensus.

This edition maps four companies turning captured CO2 and reactive mineral feedstocks into cementitious or construction-material inputs. The category is not simply about carbon capture. It is about the future supply of low-carbon SCMs: the powders, binders, and mineral ingredients that allow concrete producers to reduce clinker without rebuilding the construction industry from scratch.

Subscribe now

For years, cement decarbonization has been framed as a carbon capture problem. Capture the CO2 from kilns. Store it underground. Pay the premium. Wait for policy. Repeat.

That story is still true. But it is not the whole story.

The cement industry does not only need lower emissions. It needs new ingredients.

More specifically, it needs new sources of Supplementary Cementitious Materials — SCMs — the materials used to replace part of clinker or cement in concrete mixes. These ingredients are one of the most practical ways to reduce the carbon footprint of concrete without asking the entire construction industry to change how it builds.

For decades, the cement industry has leaned on industrial by-products such as fly ash from coal plants and blast furnace slag from steelmaking. But those legacy streams are becoming less reliable as coal plants retire, steel production shifts, and decarbonization changes the very industrial systems that used to supply the cement sector with low-carbon substitutes.

That is where the rumor starts.

Subscribe now

A new group of companies is converging around the same industrial opening: use captured CO2, alkaline minerals, kiln emissions, mine waste, steel slag, and other industrial residues to produce cementitious materials that can reduce clinker demand in cement production or replace part of cement in concrete mixes.

Not offsets. Not abstract carbon accounting. Physical inputs. Powders. Binders. Supplementary cementitious materials. New feedstocks for an industry that moves billions of tons every year and does not change its recipes lightly.

The companies we map in this edition are not identical. Some mineralize captured CO2 into stable powders. Some use CO2 to upgrade local industrial byproducts. Some screen mineral residues and design carbonation pathways around them. But they point toward the same emerging market logic: captured CO2 and reactive mineral feedstocks may become strategic inputs for the construction stack.

Together, they suggest a shift in how we should think about carbon capture, concrete, and industrial decarbonization.

Carbon may not only be stored underground.

It may be sold by the tonne into the construction supply chain.

Rumors is reserved for Premium Members.

But it’s just one part of The Scenarionist Premium experience.

If you build, back, or study Deep Tech, The Scenarionist Premium gives you the analysis, frameworks, and expert insight you need to understand where Deep Tech is really going.

Get Premium Access

1. Macro Frame: Why Now?

The cement industry has always been difficult to decarbonize because the emissions are embedded in the chemistry.

Clinker—the reactive intermediate that gives Portland cement its binding power—is produced by heating limestone and other materials to extremely high temperatures. The process releases CO2 in two ways: from fuel combustion and from the calcination of limestone itself. That makes cement different from many other industrial sectors. Even if the kiln were powered by low-carbon heat, process emissions would remain.

The International Energy Agency describes cement as “not on track” for its net-zero pathway and highlights reduction of the clinker-to-cement ratio through clinker substitutes as one of the significant levers for sector decarbonization, alongside energy efficiency, low-carbon fuels, material efficiency, and carbon capture and storage.

This is where SCMs become strategically important.

If cement producers can replace part of the clinker with lower-carbon materials while preserving compressive strength, durability, workability, and standards compliance, they can reduce emissions without waiting for a full rebuild of the cement production system.

That is the practical wedge.

Not all cement can become experimental overnight. Construction is conservative for

Read more

Every week, deep tech leaves clues.

A robot moves from demo to factory floor.
A material leaves the lab and enters qualification.
A battery company reaches a production marker.
A fusion component becomes more manufacturable.
A quantum system starts to look deployable.
A water technology meets field conditions.
A defense contract changes the financing logic around a company.
A bankruptcy reveals that the market may be real, but the capital structure was not.

Deep Tech Briefing is built around those clues.

Subscribe now

Not to celebrate every milestone. Not to chase every announcement. But to understand what each one says about the deeper transition from invention to adoption.

This week’s edition follows that transition across the full landscape: defense AI, robotics, materials, autonomy, aviation, water, carbon, geothermal, fusion, quantum, AI infrastructure, space, industrial policy, energy security, and balance-sheet stress.

The question underneath it all is simple:

What will make the difference?

Enjoy the read!

✨ Deep Tech Briefing is the weekly intelligence layer that reads the global deep tech ecosystem through the signals that matter: milestones, contracts, deployments, policy shifts, financing structures, industrial bottlenecks, and failures.

If you build, back, or study Deep Tech, it is designed to sharpen your lens — and maybe help you avoid learning a few lessons the expensive way.

Subscribe now

Anduril’s Repricing Flywheel

Contracts raise the valuation; valuation funds capacity; capacity increases the odds of winning the next contract.

There is a pattern emerging in defense technology that deserves more attention.

At first, it looks like a funding story. Anduril raised $5 billion at a $61 billion valuation, with Thrive Capital and Andreessen Horowitz leading the round. The company roughly doubled its valuation in less than a year. Defense tech has become one of the hottest categories in venture capital. Geopolitics, AI, autonomy and industrial policy are no longer separate conversations. They are becoming one market.

But the more important story is not the round itself. It is the repricing mechanism behind it.

Anduril appears to sit at the center of a new defense AI flywheel: contracts raise the

Read more

After 120 episodes, The Scenarionist is opening the platform further.

What started in 2023 as an ambitious effort to bridge the venture-building layer between science, entrepreneurship, and capital has grown into something larger: a living platform for founders, scientists, investors, operators, and technologists working at the frontier of Deep Tech company creation.

Across those conversations, we have explored the gap between scientific promise and company-building reality: fundraising, venture readiness, customer discovery, industrial scale-up, IP, team formation, go-to-market strategy, capital strategy, and the uncomfortable founder dilemmas that appear when science becomes a business.

That is why we are launching Help Me Build.

What is Help Me Build?

Help Me Build is an open Q&A space for founders building Deep Tech companies.

It is designed to answer the practical, strategic, and often uncomfortable questions that emerge between the lab, the pilot, the market, and the next financing round.

What you can ask about

You can submit questions about any part of the Deep Tech company-building journey, including:

• Venture building and company formation

• Fundraising strategy and investor readiness

• Milestone design and capital planning

• Customer discovery and commercialization

• Pilots, partnerships, and industrial deployment

• IP strategy and defensibility

• Hiring, co-founder dynamics, and team formation

• Strategic investors, non-dilutive funding, and public funding

• Scale-up, manufacturing, growth, and exit pathways

How it works

The format is simple. Through the button below, you can access a short form and submit your question anonymously.

Ask your question

I will read every submission. Then, in future issues, I will select some of the most relevant questions and answer them in a way that can help the broader community.

My hope is that Help Me Build can become a space where those questions can be surfaced more openly.

This is a beta format. I will learn by doing, and the community will help shape what it becomes.

Thank you for being part of this journey.

Stay inspired,

Nicola

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape effectively.

The Scenarionist Premium is designed to make you a better Deep Tech Founder, Investor, and Operator. Premium members gain exclusive access to unique insights, analysis, and masterclasses with the wisdom of the world’s leading Deep Tech thought leaders. Invest in yourself, and upgrade today!

Subscribe now

Dear Friends,

There is a question in Deep Tech that often appears much later than it should.

At first, it sounds like an operational question.

Something to discuss with the technical team, the COO, the supply chain people, the contract manufacturers, the equipment vendors, or whoever is responsible for turning prototypes into something that can actually be delivered.

But the more I speak with founders, investors, and operators across Deep Tech, the more convinced I become that this is not really an operational question.

It is a company-design question because, among other things, it affects:

• How much capital the company needs.

• How fast it can reach the market.

• How much margin it can capture across the value chain.

• How defensible the business becomes.

• How much control it has over quality.

• How dependent it becomes on partners.

• How credible it looks to customers.

And, consequently, the roadmap it follows and the type of company it ultimately becomes.

However, over time, markets change, supply changes, and regulation changes. That makes the manufacturing decision an evolving scenario that becomes true when the evidence catches up — a kind of Rubik’s Cube where several dimensions have to fit together within a specific window of time.

One recurring pattern across geographies and sectors is that the manufacturing conversation tends to become too binary, too quickly.

1. On one side, there is the asset-light instinct.

Stay flexible. Preserve capital. Leverage partners. Avoid building steel too early. Do not become a factory before the market is ready.

2. On the other side, there is the control instinct.

Own the process. Protect the know-how. Learn faster. Improve yield. Control quality. Do not let a third party become the place where the real manufacturing intelligence accumulates.

Both instincts are reasonable. And both can be dangerous.

A Deep Tech company that builds too early can trap itself in fixed costs before demand is real.

It can raise expensive equity to finance equipment, facilities, operators, permitting, and working capital, only to discover that customers are still testing, still qualifying, still comparing, still waiting.

But a company that outsources too casually can create a different kind of problem.

It may lose the learning loop that makes the product better. It may hand over know-how before it knows what is truly proprietary. It may become dependent on a partner that does not move at startup speed. It may struggle to control quality, cost, or yield.

This is the real pain.

Deep Tech companies are often told to scale, but not always given a clear way to decide what scaling should mean.

• They are told to be capital-efficient, but also to prove they can deliver.

• They are told to avoid CapEx, but also to show a credible path to commercial production.

• They are told to leverage partner relationships, but also to protect their moat.

• They are told to move fast, but also to build trust with customers who cannot afford supply risk.

And so the manufacturing question becomes a place where many unresolved tensions collide.

It is not just: “Should we build a factory?”

It is:

• What does the customer actually need to believe before adopting this product?

• What part of the value chain creates the real advantage?

• Is demand strong enough to justify capacity?

• Would CapEx create capacity?

• Can partners help the company move faster without absorbing the knowledge that makes it valuable?

That is the kind of decision logic I want to unpack in this piece.

Spoiler alert: There is not one answer.

A robotics company, a semiconductor company, a battery company, a bio-manufacturing company, an aerospace company, and a photonics company will not make the same manufacturing decision.

But they often face the same underlying trade-off.

They have to decide what to build, what to outsource, and what to delay.

And they have to make that decision asap, before the market, the board, the customer, or the next financing round forces the answer.

So this piece is not a manifesto for asset-light Deep Tech.

And it is not a defense of asset-heavy Deep Tech either.

It is an attempt to lay the first brick of a more useful decision language, grounded in hundreds of conversations with people doing the work in the field.

In my view, this is not a Hamlet problem of “to build or not to build.”

A better way to frame the question is:

“Which parts of manufacturing are strategic enough to own, which parts can be handled through partners, and which parts should remain flexible until the market gives the company permission to build?”

That is where the real strategy begins.

Subscribe now

The Decision Nodes We’ll Walk Through

There are many ways to approach the manufacturing question in Deep Tech. (Personally, I have identified at least 3 useful angles, and I suspect I will return to the others in future pieces…)

For this deep dive, I chose to start with these 5 decision nodes because, across the paths I have listened to and collected so far, they seem to be the most useful starting point for building a sharper lens on a decision that can shape the entire company.

For each section, I have also included practical questions and scenarios collected along the way through real-world conversations, meant to serve as a framework to help you generate sharper questions and stress-test your business plan — whether you are a founder discussing the path with your team or an investor helping a company think through the trade-offs from a different angle.

1. Product-Manufacturing-Market Fit

Read more

Deep tech is not short of ambition.

It is short of systems that can absorb ambition…

This week’s edition is about those systems. The financial system is changing as private-market exposure begins to move through new regulated channels. The defense system is shaping demand for robotics, secure communications, wireless power, and battery supply. The energy system is being pulled into AI infrastructure. The materials system is moving through qualification, pilot plants, customer testing, and project-finance evidence. The space system is becoming more investable where orbital capability maps to existing budgets.

The Big Idea gives the capital-market frame.

The rest of the briefing shows how that frame meets the industrial world.

Across robotics, defense, batteries, quantum, biopharma workflows, space, advanced materials, data centers, nuclear, water, critical minerals, and regulation, the same pattern appears: technologies are becoming valuable when they become embedded in larger systems.

That is the market we track every week.

Deep Tech Briefing is available to Premium Members of The Scenarionist.

Upgrade to read the full edition and access the intelligence layer built for deep tech founders, investors, and operators who need to understand not just what happened, but what it changes.

Unlock Premium Access

Enjoy the read!

The New LP Stack

When mutual fund holders become indirect co-investors in private markets, the cap table starts to change shape

For most of modern venture capital and private equity history, the LP base was the quiet side of the market.

The SEC staff’s recent no-action position on open-end fund participation in affiliated co-investment frameworks is not the kind of news that usually gets hype through the venture ecosystem. It does not have the spectacle of a mega-round, the simplicity of an IPO window reopening, or the mythological appeal of a founder ringing the bell at the NYSE. But I’ve been sitting with it, and I believe it is worth thinking about…

In short: for the first time, the everyday investment vehicles used by roughly 100 million Americans — such as mutual funds and ETFs — now have a path to participate in private-market co-investments alongside institutional investors. The wall between public and private capital has not disappeared. But a new door has opened inside it.

For most of modern financial history, the distinction was clean. Public markets were liquid, accessible, mark-to-market, and open to ordinary investors. Private markets were illiquid, relationship-driven, negotiated, and largely reserved for institutions, endowments, sovereign funds, family offices, and accredited individuals. One was distribution. The other was access.

That distinction is now beginning to blur.

We are entering the era of

Read more

Subscribe now

This week, I sat down with Roman Wolff, Industry Expert in Manufacturing Scale-Up, to unpack how an experienced manufacturing and engineering operator thinks about moving Deep Tech companies from the lab to industrial scale.

Why Scale-Up Is Not Just a Bigger Lab

In Deep Tech manufacturing, the term scale-up is often used as soon as a company moves from gram-scale experiments to kilogram-scale production.

That transition is meaningful. It allows the team to handle more material, generate larger samples for customers, and learn how the process behaves outside the smallest laboratory conditions.

From an engineering and manufacturing perspective, however, this stage is still often closer to an expanded lab environment than to true industrial scale-up.

The distinction matters because producing more material does not automatically mean that the company is learning what it needs to know to design, finance, and operate a commercial plant.

From grams and kilos to real manufacturing questions

At the early stage, the central question is usually whether the chemistry works. The company is focused on proving the core technical principle, generating data, and producing enough material for evaluation.

As the process moves toward industrial relevance, the questions become more practical and more operational.

The company has to understand the role of recycle streams. It has to evaluate reliability and determine how long equipment can run under realistic conditions. It has to consider materials of construction. It has to assess product stability, shelf life, handling, and storage.

These are the kinds of considerations that determine whether a process can become manufacturable.

They are also the factors that may remain invisible when the company is working only with gram or kilo quantities in a laboratory setting.

In that environment, many streams can be handled manually. Small losses may not matter. Equipment uptime may not be a central concern. Materials compatibility may not yet create a constraint. The process can look successful while still being far from industrially proven.

The shift toward scale-up begins when those hidden issues start to surface.

Why scale is defined by equipment, not size

A useful way to think about scale-up is that it is defined less by volume and more by equipment similarity.

The important question is not only how much product the company can make. It is whether the company is using equipment that can provide relevant information for the future plant.

This is why engagement with equipment suppliers becomes an important milestone.

Vendors understand the unit operations, equipment limits, sizing logic, and performance guarantees associated with their systems.

Their input helps the company understand what is commercially available, what the smallest scalable unit might be, and what assumptions need to be tested before larger capital decisions are made.

This also helps reduce equipment risk.

In a first-of-a-kind (FOAK) plant, the technology may already introduce enough uncertainty. The chemistry may be new, the process integration may be new, and the commercial model may still be developing. Adding custom or unproven equipment can increase the risk profile significantly.

The preferred approach is to keep the newness concentrated in the chemistry or process innovation while relying, as much as possible, on standard equipment.

That does not remove the need for engineering work. Standard equipment still has to be selected, configured, tested, and integrated into the broader process. But it gives the company a more credible path toward scale because the equipment has an industrial reference point and a supplier who can support it.

The first signs that a process is becoming industrial

A process starts becoming industrial when the team begins to describe the full plant, not only the core chemistry.

That includes the feedstock coming in, the product going out, and every intermediate stream that has to be handled along the way.

A practical way to frame the problem is to ask what a truck will bring into the site and what a truck will take away from it.

That question changes the level of analysis.

It forces the team to define feedstock specifications, product specifications, byproducts, recycle streams, waste streams, storage requirements, and logistics.

It also reveals practical issues that may have been invisible in lab work: how to load a catalyst, how to separate solids from liquids, how to move material between steps, how to manage recycles, and how to keep the process running reliably.

These details can look secondary from a research perspective. From a manufacturing perspective, they are central. They influence the process design, the cost estimate, the pilot strategy, the vendor conversations, and ultimately the investment case.

A company becomes more credible when it can show that it understands this transition.

The move from lab work to scale-up is therefore less about declaring that more material has been produced and more about demonstrating that the process is being translated into an industrial system.

Designing the Scale-Up Path Backward from the Commercial Plant

A practical scale-up roadmap starts from the end state.

Before deciding what the next pilot should look like, the company needs to form an initial view of the commercial plant it is eventually trying to build.

That view will still be incomplete and uncertain, but it gives structure to the work. It allows the team to understand which questions must be answered, which equipment must be tested, and which assumptions will drive the techno-economic model.

At this stage, the company may already have a product concept, a potential customer, and some form of commercial signal. A customer may have indicated interest in buying a defined volume at a certain price, or may have provided enough information to begin estimating demand. That creates the basis for early economics.

The first task is to turn that commercial signal into a manufacturing problem.

The company has to understand what product specification it must reach, what feedstock it can actually buy, and what the plant must do in order to convert one into the other. The result is an early picture of the commercial facility, built backward from the market requirement.

Starting with the product specification and the feedstock specification

The product specification is one of the first anchors of the scale-up process.

If there is a real customer or a credible market pull, the customer defines what the final product must be. The product has to meet a specification, and that specification becomes the target for the process design.

At the same time, the company has to look carefully at the feedstock.

In the lab, materials are often purchased in forms that are convenient for research. They may be high purity, available in small quantities, or selected because they allow the chemistry to be tested cleanly.

A commercial plant operates under a different logic. The feedstock has to be something the company can actually purchase at scale, with real specifications, real availability, and real supply constraints.

This is an important shift in thinking.

The company is no longer asking only whether the chemistry works with ideal inputs. It is asking whether the process can work with materials that can be supplied to an industrial facility. That feedstock specification becomes part of the economic and technical model.

If the feedstock is unrealistic, the rest of the analysis becomes fragile. If the product specification is unclear, the process has no stable endpoint.

Both sides of the equation need to be defined early enough to guide the design.

Mapping each step of the future plant before building the pilot

Once the company has an initial view of the commercial plant, each step in the process needs to be assigned to a type of equipment.

At the beginning, there may be more than one possible equipment choice for a given step. That is expected at an early stage. The important point is that every step is being connected to an industrial unit operation.

This process also defines the scale-up path.

When the company speaks with equipment suppliers, the vendors can help identify the commercial-scale equipment and the smallest scalable version that can be purchased for pilot work.

That smallest scalable unit often determines the size and structure of the pilot.

The design should be guided by the questions that must be answered for the commercial plant.

• If a certain equipment type is likely to be used at commercial scale, the setup should generate the data needed to validate that choice.

• If recycles will matter in the plant, the pilot should help the company understand them.

• If reliability, stability, or materials of construction are important, the pilot should create conditions where those questions can be studied.

This is also where early techno-economic model begins to become more grounded, and it becomes a working tool for improving the design.

At this stage, the company may have enough information to prepare a funding proposal for the next phase of work.

That funding is usually meant to support the pilot, the first level of engineering work, and the company’s operating runway while those activities are being completed.

This is still an early investment case. It is not yet the funding case for the full commercial plant. There are too many uncertainties. The process still needs to be tested, the design still needs to mature, and the risks need to be reduced step by step.

How FEL Turns Technology into an Investment Case

Front-end loading, often described through FEL1, FEL2, and FEL3, is the process that progressively turns a technical concept into a Final Investment Decision (FID).

In the early stages of a Deep Tech company, many parts of the industrial plan are still fluid. The chemistry may work. The product may have customer interest. The company may have an initial view of the commercial plant.

However, the level of engineering definition is still too low to support a major capital decision.

The FEL process creates structure around that uncertainty.

Each phase increases the maturity of the design, reduces the range of cost uncertainty, and forces the company to make decisions that were previously open.

The work moves from a high-level view of the process to a detailed definition of what will be built, where it will be built, which equipment will be used, and how much the project is expected to cost.

This matters because the investment case has to show that the process has been translated into a defined project.

FEL1: fixing the process logic and learning while it is still cheap

FEL1 is the stage where the company fixes the basic process logic. At this point, the block flow diagram is defined.

The company knows the main steps required to move from feedstock to product. It may still be converging on the exact scope, and many details remain open, but the overall route is becoming clear.

The cost estimate is still relatively broad because the design is still early. The purpose of FEL1 is not to create final precision. It is to define the process architecture well enough to understand what must be studied, tested, and improved.

This is also one of the best stages for learning. Changes are still relatively inexpensive.

The company can update assumptions, improve the flow diagram, revisit alternatives, and use feedback from vendors, engineering partners, and pilot work to strengthen the design.

The goal is to create enough clarity to move into the next phase with a more disciplined scope.

For Deep Tech companies, this phase should also interact closely with the pilot strategy. FEL1 and pilot planning can inform each other. As the company learns more from equipment suppliers and early testing, the engineering view can be refined.

That cycle is valuable because it helps identify problems before they become embedded in a more mature design.

FEL2: converging on scope and completing the technology work

FEL2 is where the process moves from block flow logic to a more defined process flow diagram.

At this stage, the company is no longer only defining the major steps. It is beginning to decide how those steps will be performed.

This is where scope convergence becomes critical. The company needs to define the type of equipment that will perform that operation.

FEL2 is also the stage where alternatives are reviewed and major choices are narrowed.

The company should be locking in site selection, or at least moving toward a level of site definition that allows the next phase of design to proceed. It should also be completing the pilot and technology development work that is required for the project.

This point is especially important.

By the end of FEL2, the company should not expect to continue fundamental technology development for that specific plant design. The pilot should have reduced the key technical risks. The process data needed for the project should be available.

The major questions that could change the design should have been answered.

FEL2 therefore acts as a gate between technology development and detailed project definition. It is where the company proves that the process is mature enough to support a more precise engineering effort.

FEL3: locking the design before the Final Investment Decision (FID)

FEL3 is the stage where the project becomes highly defined. At this point, the company should know what it is building.

The work is focused on the details required to support execution.

This includes piping and instrumentation diagrams, plot plans, 3D models, vendor selections, and vendor quotes. The scope is fully locked. The design has moved from a conceptual or semi-defined process into a project that can be costed, reviewed, and prepared for execution.

At FEL3, the company should not be using engineering to solve unresolved technology questions. It should be using engineering to finalize a project whose technology basis has already been validated.

That distinction is central to capital efficiency.

Late-stage design changes are expensive. If a company discovers during FEL3 that the pilot data is insufficient, the site assumptions are unstable, or a core unit operation has not been proven, the schedule and budget can be significantly affected.

The earlier stages exist to prevent that situation.

FEL3 creates the technical and cost basis required for the investment decision. It gives the stakeholders enough definition to understand the scope, the capital requirement, the project risks, and the path to construction.

The final investment decision (FID)

The final investment decision is often discussed alongside the engineering process, but it is fundamentally a financial milestone.

Engineering creates the conditions for FID. It provides the defined scope, mature cost estimate, vendor support, site information, and project design required to make a capital decision. However, the decision itself is about committing capital.

At FID, the company or its financial backers decide whether to fund the project and move into execution.

For a large industrial plant, that may involve not only the direct capital cost of construction, but also fees, interest, company operating costs during the construction period, and the capital needed to support the organization until the plant is producing.

This is why the quality of the FEL process matters so much. For a Deep Tech company, this is the point where technical progress and financial readiness begin to converge.

The Real Timeline of Industrial Scale-Up

Industrial scale-up requires a timeline that is often longer than early commercial discussions suggest.

Once a company has customer interest, an initial product specification, and a view of the process, it can be tempting to think of the next step as a direct move toward production.

In practice, the path from early scale-up planning to a commercial plant includes several parallel workstreams: engineering, piloting, site selection, permitting, financing, construction, hiring, and commissioning.

The duration depends on the size and complexity of the project. A lower-capex project may move faster. A larger first-of-a-kind plant will usually require more time, more definition, and more capital discipline.

For a large industrial project, the path to mechanical completion can easily extend across 4 to 5 years if the sequence runs well.

That timeline includes the engineering phases, the final investment decision, detailed engineering, construction, and the early work required to prepare the organization that will operate the plant.

Why a commercial plant can take 4 to 5 years to reach mechanical completion

The front-end engineering process alone can take more than a year and a half for a large project.

FEL1 may take 3 or 4 months. FEL2 may take roughly 6 months. FEL3 may take around 12 months, assuming the required information is available and the project does not have major unresolved issues.

Those timeframes are indicative, and they vary by project, but they give a realistic sense of the planning horizon.

After FEL3, the company reaches the point where it can move toward a final investment decision. That decision can take additional time, potentially several months, because it involves financing, risk review, commercial agreements, and the final commitment of capital.

Once the investment decision is made, detailed engineering and construction begin.

For a large project, that phase can take approximately 2 years, assuming there are no unusual constraints around long-lead equipment or other major delivery issues.

Taken together, the industrial timeline becomes substantial.

The company may spend around 18 months or more reaching the end of FEL3, additional time moving through the FID, and roughly 24 months on detailed engineering and construction.

This is how a project can become a four-to-five-year journey from the decision to pursue scale-up to mechanical completion.

How permitting, site selection, engineering, and construction move together

Several activities have to move in parallel, and delays in one area can affect the rest of the timeline.

1. Site selection is one of the most important examples. The company needs a defined site to complete the design properly. Without a site, many engineering assumptions remain unstable. Layout, utilities, logistics, permitting requirements, and other site-specific design assumptions all depend on where the plant will be built.

2. Permitting also affects the schedule. A company typically needs enough engineering definition before it can apply for permits. That means permitting often begins around FEL3, when the project has enough detail to support the application. The permitting process then runs alongside the later project work and should ideally be sufficiently advanced by the time the company is ready to proceed after the FID.

3. Construction timing also depends on the equipment profile. If the project requires equipment with long delivery times, the schedule may extend. If there are unusual components, custom systems, or supply chain constraints, the construction plan has to absorb those realities. In a FOAK plant, these dependencies need to be managed carefully because the project is already carrying technology and execution risk.

The overall timeline therefore has to be treated as an integrated plan.

Engineering is connected to site selection. Site selection is connected to permitting. Permitting is connected to the investment decision. The investment decision is connected to procurement and construction. Each phase creates the conditions for the next one.

Why first product and design-rate production are different milestones

Mechanical completion does not mean the plant is already producing at full commercial performance.

Once construction is complete, the company still has to start up the plant, commission the systems, bring materials into the facility, and begin operating the process under real conditions.

For FOAK technologies, this ramp-up period is particularly important because unexpected issues often appear only when the plant is running.

The first product may come relatively soon after startup. In a successful case, the first truck leaving the plant could happen perhaps a month after startup.

However, reaching design-rate production can take longer.

A reasonable planning assumption may be 3 to 6 months after mechanical completion, depending on the complexity of the process and the issues that appear during commissioning and early operations.

This distinction matters because early production and stable production are different operating states.

A plant may demonstrate that it can reach the intended design rate early, but then experience equipment failures or reliability problems. These issues can be highly specific.

Gaskets may leak because the selected material was wrong. Certain components may behave differently under continuous operation than they did in the pilot.

These are normal learning points in early plant operation.

The pilot reduces risk, but it cannot reveal every issue that will appear in the full-scale facility. Some problems only emerge when equipment is connected at scale, materials are moving continuously, and the plant is operating as an integrated system.

This is why the timeline should include a realistic ramp-up period.

The company needs time to troubleshoot, modify, repair, optimize, and stabilize the plant. From an investment and customer perspective, the relevant milestone is not only first production. It is the ability to operate reliably and approach the intended production rate.

For Deep Tech companies, this has direct implications for cash planning.

The company needs enough capital not only to design and build the plant, but also to operate through commissioning and ramp-up. The period between mechanical completion and stable production can still consume cash. If the plan assumes immediate full-rate production, the financing model may become too optimistic.

A realistic industrial roadmap therefore separates the major milestones clearly: engineering completion, FID, mechanical completion, first product, and design-rate production.

Each milestone represents a different level of readiness, and each requires its own assumptions about time, capital, and execution risk.

Building the Operating Company Before the Plant Starts

The construction phase is not only a period for building the physical plant. It is also the period in which the company has to build the organization that will run it.

By the time mechanical completion arrives, the plant cannot be treated as a finished asset waiting for a team to appear around it.

The company needs the people, systems, procedures, supply chain, logistics, and operating structure already in place. Otherwise, the project may reach physical completion without being ready to operate.

This is one of the most important differences between a technology company preparing for scale-up and an industrial company preparing for production.

A plant is not only equipment. It is an operating system.

Why offtake agreements matter before the investment decision

Before the final investment decision, the company should ideally have its offtake agreements in place.

This is especially important for large capital projects. If a project requires hundreds of millions, or even more than a billion dollars, the investors or lenders backing the plant will want to know who is buying the output, under what terms, and for how long.

A potential buyer’s interest is useful, but it is usually not enough to support a major plant investment. The financial backer will likely want a committed customer, often under strong contractual terms, for a meaningful share of the plant’s production.

That agreement becomes part of the financing logic.

It shows that the project is connected to demand. It reduces uncertainty around revenue. It gives investors more confidence that the plant, once completed, will have a market for its output.

Hiring the manufacturing leader early

The operating organization should begin forming before the plant is finished.

One of the first important hires is the manufacturing leader, whether that role is called site director, plant manager, or another equivalent title.

This person should join early enough to influence the engineering and design process, not only to manage the facility after construction.

That timing matters.

An experienced manufacturing leader can provide practical input that improves the plant design. They can identify operational issues, maintenance concerns, staffing needs, safety considerations, commissioning requirements, and design choices that may look acceptable on paper but create problems in day-to-day operation.

This is particularly valuable in a FOAK plant.

Engineering teams can design the system, but operating leaders understand what it takes to run the system continuously. They bring the perspective of reliability, maintainability, procedures, staffing, and plant discipline.

Hiring this person early also creates a bridge between the project team and the future operating team. The plant is then being designed not only as a capital project, but as a facility that people will have to operate safely and reliably.

Why the ramp-up exposes what the pilot could not reveal

Even with a strong pilot, good engineering, experienced vendors, and a capable operating team, the first commercial plant will still teach the company new lessons.

FOAK technologies tend to fail in unexpected ways. The pilot can reduce risk, but it cannot perfectly reproduce all the conditions of the full plant.

When the equipment is larger, the streams are connected, the process runs continuously, and real production targets apply, new issues can appear.

Some of those issues may be mechanical. Other issues may involve reliability, maintenance, startup procedures, operator training, or supply logistics.

This is why ramp-up should be treated as part of the scale-up plan.

The company should expect a period of troubleshooting and stabilization after mechanical completion. First product may come relatively early, but stable production at design rate takes more time.

The operating company therefore has to be ready before the plant starts.

It needs the people who can identify problems, fix them, document them, and keep improving the system. It needs the procedures and supply chain needed to support continuous operation. It needs enough capital runway to move through commissioning and early production without assuming that everything will work perfectly on day one.

The broader lesson is that scale-up is not only an engineering sequence. It is an organizational buildout.

The company has to mature at the same time as the plant. By the time the facility is mechanically complete, the business must already have become capable of operating it.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape effectively.

The Scenarionist Premium is designed to make you a better Deep Tech Founder, Investor, and Operator. Premium members gain exclusive access to unique insights, analysis, and masterclasses with the wisdom of the world’s leading Deep Tech thought leaders. Invest in yourself, and upgrade today!

Subscribe now

Dear Friends,

I’m back with a new piece based on lessons learned.

This time, we will start from an often overlooked part of the journey.

The end.

When I spend time every week with Deep Tech founders, investors, and operators, I notice that the most valuable conversations do not start with answers.

They start with better questions.

And over the past few months, one question has kept coming back from different angles: “When should a Deep Tech company really start thinking about its exit strategy?”

The common answer is: later.

• After the product works.

• After the first customers.

• After the next round.

• After the company is more mature.

• After the market understands what is being built.

These answers sound reasonable.

After all, Deep Tech is hard because getting started is hard. Moving from the lab to the real world is hard. Convincing customers to trust something new is hard. Raising capital before the market fully understands what you are building is hard.

But the more I listen to people who have actually built, backed, acquired, or helped Deep Tech companies reach the public markets, the more convinced I become that, in Deep Tech, exit is rarely just the final part of the conversation.

It often starts shaping the company much earlier than founders realize.

The reason is simple. It can affect:

• which customers you prioritize

• which investors you bring in

• how you structure pilots

• how you protect IP

• how you think about manufacturing

• how much capital you raise

• what kind of strategic partnerships you accept

• and, of course, whether the business you are building can eventually become legible to an acquirer or the public markets.

That is what makes this topic so important.

However, in Deep Tech, exit strategy is often discussed too late. It is treated as the polite answer to a predictable investor question. Something to revisit once the company is “ready.”

But in many companies, exit readiness is not created at the end. It is defined through decisions made much earlier.

It is the pivotal step where scenario planning and a backcasting approach help founders start from the end — or, even better, from several possible futures — and then draw a line back to the present.

In my view, this is the core of an exit strategy.

So, I’ve invested plenty of time trying to understand what really holds across sectors, technologies, markets, and geographies.

What is just local noise. What is sector-specific. What depends on timing. What depends on capital markets. What depends on luck. And, over time, some of those patterns begin to converge.

• A few strategic mistakes keep repeating.

• A few design approaches keep standing out.

• A few assumptions keep separating companies that look promising from companies that can actually become exit-ready.

That is what pushed me to write this piece.

These are 30 exit strategy lessons I have been collecting and refining through conversations with Deep Tech founders, investors, and operators — first because I wanted to sharpen my own understanding of how Deep Tech companies become valuable assets, and then because I realized these lessons might be useful to others building in the same reality.

• Some lessons are about setting the North Star strategically.

• Others are more tactical, focused on how to navigate the road when the exit path is already becoming real.

Taken as a whole, they are meant to help founders and investors think more clearly about how exit logic shows up long before the exit itself.

Together, we’ll explore why the most successful founders understand who might need the company one day, why that buyer would act, what capital structure preserves optionality, and what kind of business the market can eventually underwrite.

As always, the result is not a definitive map of every possible outcome.

It is a condensed one. And, I hope, a useful one.

If you build, back, or study Deep Tech, I think these lessons will help you ask sharper questions about exit strategy design — and maybe avoid discovering too late that the company you built is not as easy to back, acquire, underwrite, or integrate as you once assumed.

Subscribe now

Lesson 1

A signed term sheet changes the buyer’s clock.

The next financing round can turn strategic interest into acquisition urgency.

Lesson 2

The business model must fit the exit strategy.

A royalty stream, a factory, and a venture-scale platform do not exit the same way.

Lesson 3

Set the IPO as the North Star, but keep the off-ramps visible.

The best companies build toward scale without pretending there is only one ending.

Lesson 4

Read more

Deep tech is increasingly being shaped by systems outside the startup itself...

The grid determines whether compute, electrolysers, factories, and advanced manufacturing can expand. Defence procurement determines which AI and robotics companies become embedded in national security infrastructure. Sovereign capital determines which technologies are treated as strategic assets. Regulation determines whether clinical trials, digital markets, supply chains, and export controls accelerate or constrain entire categories.

This week, all of those systems moved at once.

But the clearest signal came from one extraordinary financing event: Ineffable Intelligence raised $1.1 billion in a seed round at a $5.1 billion valuation.

The number matters, of course... But there is much more to think about when a seed round can exceed the committed capital of many European venture capital vehicles. That is what we discussed in this week’s The Big Idea.

Because this round is not only about one company, one founder, or one AI thesis. It is about what happens when capital starts underwriting systems that may learn through action rather than human-generated data alone; and when the defensible asset may move from the model to the simulator, the environment, and the feedback loop.

Then, in The Week in Milestones, which, as every week, is our observatory on the key milestones changing company trajectories, we move from capital formation to vertical integration, from commercial proof to manufacturing scale-up, and from qualification gates to technical progress across defence, industrial software, construction automation, biomanufacturing, robotics, batteries, quantum, fusion, geothermal, solar, and advanced materials.

And looking at the bigger forces rewriting the conditions for building, this week we analyze the electricity bottleneck behind AI, data centres, and advanced manufacturing; the integration of frontier AI into classified defence systems; China’s new supply-chain security regulations; the rise of sovereign capital as an industrial policy tool; and the regulatory shifts that could reshape clinical trials, digital markets, export controls, and supply-chain strategy.

The market is no longer just funding technology.
It is funding control over the systems that make technology usable.

Deep Tech Briefing is available to Premium Members of The Scenarionist.

Upgrade to read the full edition and access the intelligence layer built for deep tech founders, investors, and operators who need to understand not just what happened, but what it changes.

Unlock Premium Access

Enjoy the read!

The $1.1 Billion Bet on the End of Human-Generated Data

What It Means When an AI Seed Round Weighs More Than a European Venture Fund

Let’s be precise about what happened on April 27. A five-month-old company with no public product, no disclosed revenue, or commercial timeline raised $1.1 billion in a seed round at a $5.1 billion valuation.

Sequoia flew Alfred Lin and Sonya Huang to London personally to secure the deal, with Lightspeed Venture Partners co-leading the round. NVIDIA participated, with reports suggesting a check of at least $250 million. Google, whose DeepMind unit remains one of the strongest institutions in reinforcement learning, co-invested. The UK government showed up with sovereign capital. The venture world called it the largest seed round in European history.

This is the story of Ineffable Intelligence.

And here is the point: an initial financing round that can exceed the committed capital of many European venture capital vehicles cannot be treated as a normal startup event. It changes the meaning of early-stage financing, and it raises a set of questions that should be taken seriously…

Read more