Welcome to the 127th edition of Deep Tech Catalyst, the educational channel from The Scenarionist where science meets venture!
This week, I sat down with Sean Mitchell, co-founder and former CEO of Movidius, the fabless semiconductor company acquired by Intel in 2016. Drawing on his founder journey, we explored what it takes to navigate market uncertainty, customer evidence, capital constraints, and timing when building in complex semiconductor markets.
Key takeaways from the episode:
🎯 The first market is a wedge
A beachhead gives a company focus, but it should not become a constraint. Early customer conversations often reveal that the technology matters, just not in the way the founders initially packaged it. The best founders know how to separate rejection of a first use case from rejection of the technology itself.
🧪 Show proof before asking the market to believe
In semiconductors, customer discovery rarely starts with a blank conversation. Serious feedback comes when a startup can put something concrete in front of the market: a benchmark, simulation, test chip, prototype, or technical demonstration. Proof changes the conversation from abstract interest to specific technical evaluation.
🤝 Turn scarcity into customer qualification
For a resource-constrained deep tech startup, early product access is one of the few sources of leverage. The right strategic customers are not simply the biggest logos, but the ones willing to commit engineering time, define concrete requirements, and help de-risk the next product or financing milestone.
⏳ Convert momentum into runway
In deep tech, survival is a sequencing problem. Technical proof must become customer evidence, customer evidence must become capital, and capital must become more time. Staying in the game is not about optimism; it is the operating discipline of extending the company’s runway until technology, market, and timing finally align.
The Humanoid SPAC Test | Deep Tech Briefing 117
Independent intelligence for deep tech allocation and industrial strategy.
BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE
Start With a Technical Conviction
In 2005, the Movidius founding team was not simply looking for a startup idea in semiconductors. It was looking at a structural inefficiency in computing.
A growing class of workloads was poorly served by existing architectures. Traditional processors could execute instructions quickly, but for these workloads, much of the system’s effort was consumed moving data rather than performing useful computation.
That distinction mattered. The opportunity was to ask whether a different architecture could make these workloads dramatically more efficient.
This was the original conviction behind Movidius: emerging applications would require a more efficient way to process data-intensive workloads, especially as these workloads moved toward constrained computing environments such as mobile devices.
The founders believed that a specialized architecture could lift performance dramatically by reducing the inefficiencies embedded in general-purpose computing.
At the time, however, the market was not yet obvious.
This is an uncomfortable but common starting point in deep tech.
In semiconductors, advanced materials, robotics, energy systems, and other frontier technologies, the sequence is often less linear. The technical capability may precede the market category.
The founder may see that a new capability is becoming possible before customers have a precise language for how they will buy it, integrate it, or prioritize it.
It means the team has to manage two different forms of uncertainty at the same time.
Can the team actually build what it believes is possible?
Where will this capability first become urgent enough to justify adoption, and in what product form?
That is where strategic discipline begins: knowing which part of the company’s logic must remain stable and which part must remain open.
A founder who becomes too attached to the first market hypothesis may force a powerful technology into the wrong application.
A founder who treats the technical thesis as infinitely flexible may lose the very insight that made the company worth building in the first place.
For Movidius, the stable element was the belief that data-intensive computation needed a more efficient architecture. The open question was the entry point. The first commercial application had to be concrete enough to test the company’s assumptions against real customers, real constraints, and real purchasing behavior.
This is a subtle but important lesson for deep tech builders. The first market map is rarely the final one. Its purpose is to create a disciplined path into reality and to clarify what must become true in the world for a technical enabler to become a solution customers urgently need.
Proof Before Product
In deep tech, customer discovery often fails because it is treated as an interview process rather than a proof-building process. A founder approaches a large customer and asks: “What problems do you have?”
The question sounds customer-centric, but in practice it usually produces incomplete answers.
The customer may not yet have a concrete way to map an emerging capability onto its own problems.
The people in the room may not own the most important problem. Or the real constraint may sit somewhere between R&D teams, product teams, executives, and other decision-makers up the organization.
This is especially true in semiconductors.
A chip company cannot simply ask the market what to build and expect a clean answer. The development cycles are too long, the integration risk is too high, and the product-definition stakes are too large.
Customers need to see enough evidence to believe the architecture is credible before they will invest the time required to explain what they truly need.
The way through is to bring proof before product.
That proof can take different forms: a simulation, a prototype, a test chip, a performance target, or a technical demonstration that shows a credible piece of the future system. It does not need to be the final commercial product. It does need to be concrete enough to change the quality of the conversation.
The company’s early experience reflected this pattern.
For more than a year, the team worked through a technology feasibility phase before customer engagement became very active. As the architecture started to show promise, the founders could approach potential customers not with a blank page, but with evidence.
They could show a direction of travel and make the technical claim concrete enough to be tested in discussion. They could give sophisticated technical buyers something specific to evaluate.
That changed the customer conversation.
Instead of asking customers to imagine an abstract capability, the team could frame a more useful discussion: this is what the architecture appears able to do; this is the kind of performance it can deliver; this is not yet the final product, but it suggests a path.
The difference is subtle but decisive. Customers often cannot describe the future in the abstract. But they can react intelligently to a credible, tangible proof point.
They can say whether a performance level is meaningful, whether an interface is missing, whether integration would be painful, or whether a use case is commercially relevant.
In that sense, early technical proof is a customer discovery tool. It also helps reveal which customers are merely curious and which customers are serious.
Serious customers will give a challenge. They will define a performance target. They will describe a specific workload. They will expose a constraint that must be solved before adoption becomes possible.
For founders, this is one of the most valuable signals in the early commercial process. A precise technical challenge is much more useful than a simple expression of interest. It means the customer has moved from passive interest to active evaluation.
Some of the strongest conversations in the company’s journey emerged when customers could articulate the problem clearly, saying something along these lines: “If the technology can run this workload at this level of performance, then we are interested.”
That kind of statement gives the team a target, a reason to allocate engineering resources, and eventually a form of evidence that investors can understand.
This matters because in deep tech, market risk and technical risk are intertwined.
A customer commitment is only meaningful if the technical requirement is real. A technical milestone is only valuable if it maps to a customer problem that matters. The founder’s task is to connect the two as early as possible.
For semiconductor founders, the lesson is particularly important because the final product may take years, and integration into a customer system may depend on requirements that are easy to miss and expensive to correct. The goal is to convert uncertainty into sharper questions before the most expensive decisions are locked in.
Commitment Is the Signal
An early-stage company needs credibility, revenue, and reference points. But it also has real constraints. Those constraints create scarcity. Used well, scarcity becomes a way to qualify seriousness.
One of the strongest lessons from this case study is that early access to the technology could not be offered casually to every interested customer. It had to be treated as a selective resource.
The company could only support a small number of early adopters, so the question became whether a customer was serious enough to justify that scarce capacity.
The message was no longer simply about inviting customers to try the technology. It became a test of intent: if a customer wanted early or preferential access, the startup needed evidence that the opportunity mattered, that there was a real project behind the interest, and that successful collaboration could lead to meaningful business.
That distinction matters. A letter of intent may be useful in the earliest phase, when the company is still trying to organize market conversations and create basic evidence of interest.
But as the company matures, the stronger signal is when a customer commits something scarce of its own: money, defined volumes, upfront commitments, or a concrete mechanism tied to access, capacity, pricing, or future delivery.
This is where early customer engagement becomes strategic.
The company is building a chain of evidence. A serious customer commitment can help justify an expensive product decision. It can give investors confidence that the market signal is real. It can sharpen technical priorities. It can also create urgency inside the customer organization.
This is why the best early adopters are the customers whose urgency is specific, whose technical requirements are clear, and whose willingness to commit is real. They help define the next proof point.
For a semiconductor company, that distinction is especially important. The move from prototype to production can require expensive decisions, manufacturing-related commitments, packaging and testing work, and long development timelines.
Before making those investments, the company needs a reason to believe that a defined product, built to defined specifications, will matter enough for customers to adopt. A committed early customer can provide that reason.
Product Definition Is a Journey of Precision
Product definition is rarely a single decision made once at the beginning. It is a process of increasing precision. A company may start with a powerful technical conviction: that a new architecture can solve a class of problems existing systems are poorly equipped to handle.
But a technical conviction does not become a company until it finds the right commercial form.
That form is discovered over time, through proof points, customer conversations, integration constraints, capital pressure, and the discipline to keep moving before the answer is fully visible.
This journey brings that lesson into sharp relief.
The team moved through successive stages of learning about which applications mattered, which customers felt the problem most sharply, and which product definition could turn architectural advantage into commercial adoption.
This is especially true in semiconductors because the product does not live alone.
A chip has to fit inside another company’s system: its device architecture, engineering process, cost structure, roadmap, and internal decision-making logic.
Performance matters, but performance by itself is not enough. Adoption depends on whether the technology can become usable inside the customer’s world.
In the case discussed, the company did not arrive at the winning product definition in a straight line. It advanced through stages: feasibility work, early customer engagement, application learning, sharper technical challenges, and eventually customer evidence strong enough to support the next major product direction.
Each stage narrowed the distance between what the architecture could do and where the market most needed it.
The founder must convert belief into evidence, evidence into product direction, and product direction into enough credibility to earn more time.
Staying in the game does not mean waiting passively for the market to catch up. It means learning fast enough, and focusing sharply enough, to keep the company alive while the right opportunity takes shape.
In this case, the decisive turn came around the second major product version, when a concrete customer challenge changed the trajectory of the company.
For founders, the lesson is that each product step should make the company smarter. Each prototype, customer exchange, and technical proof point should reduce uncertainty and sharpen the next decision.
The winning product may not be obvious on day one. It may emerge only after the company has tested entry points, earned customer intimacy, survived capital constraints, and held the technical vision long enough for the market to reveal the right wedge.
That is why perseverance matters: the ability to keep the core conviction intact while allowing the product to become more precise.
The Founder’s Real Job Is to Stay in the Game
Successful companies compress time. From the outside, a company’s journey becomes a clean narrative: a technical insight, a few strategic pivots, a breakthrough product.
Often, the story appears coherent because the ending gives structure to everything that came before it. But that is not how the journey feels from inside the company.
Inside, the path is usually discontinuous. Progress comes in fragments. A test chip works, but the product definition is incomplete. Customers are interested, but not yet committed. Investors are supportive, until the macro environment changes.
The next product looks like the right one, but the company may not have enough capital to reach it.
This is why the central operating problem in deep tech is not only whether the technology can work. It is whether the company can survive long enough for the technology, the market, and timing to converge.
The company’s journey makes that reality unusually clear when, after closing its first institutional round, the team entered a dramatically different funding environment.
Capital became harder to access. Risk appetite narrowed. A semiconductor startup, already facing long development cycles and expensive product milestones, suddenly had to operate under a much harsher assumption: it might not be able to raise again for some time.
That changed the meaning of every decision.
Spending discipline was no longer an operational preference. It became a survival requirement. The company had to focus capital on what mattered most: building the product, selling the product, and creating enough evidence to justify the next step.
Anything outside that core had to be questioned.
This is one of the most important lessons learned along this incredible journey. Vision may attract the first believers, but survival depends on sequencing. The founder must constantly ask:
What is the next proof point that unlocks more time?
Sometimes that proof point is technical. A feasibility study. A working prototype. A test chip. A performance demonstration that shows the architecture can deliver what the market requires.
Sometimes it is commercial. A serious customer conversation. A defined workload. A performance target. An upfront commitment. Sometimes it is financial. A new investor willing to underwrite the next development cycle because customer evidence has made the risk more legible.
The founder’s job is to connect these proof points into a path.
That path is rarely elegant. In this case, it included reducing burn, closing offices, laying off people, and facing pressure to sell the company before the founders believed the real opportunity had been built.
The founder’s responsibility is to keep the company moving through that sequence. Not indefinitely. Not blindly. But long enough for the right evidence to appear—and disciplined enough to recognize it when it does.
This is the kind of moment that rarely appears in retrospective success stories with the force it deserves.
