The ten-year founder in a five-year VC ecosystem

Most published advice on building startups is written for software founders. Much of it is practically useless in DeepTech. The dominant vocabulary, built on the seminal works of Eric Ries's Lean Startup theory and Steve Blank's Customer Development methodology, was created for software businesses that can build, ship, learn and iterate at speed.

When founders are told to "trust the process", it is this loop of rapid iteration they are being told to trust. The frameworks are excellent in their native context, and they have become foundational to how an entire generation of founders thinks about building a company.

So deeply ingrained have these frameworks become in the startup psyche that they have almost become invisible - their heritage forgotten. Founders absorb the assumptions without noticing they are assumptions, and the venture ecosystem has organised itself around them: the advice columns, the accelerator curricula, the questions investors ask in diligence.

For a DeepTech founder, this creates a problem that is hard to name precisely because it is everywhere. The SaaS rapid iteration loop is the wrong shape for the work. What follows is an exploration of two failure modes it produces, one internal, one external, and what the strongest DeepTech founders do about each.

The signal drought

DeepTech feedback cycles are measured in years, not months. An advanced materials founder may only achieve a meaningful technical milestone once every six months and a substantive commercial conversation even more rarely in the early stages. The psychological problem this creates is not the absence of progress; it is the absence of evidence of progress that anyone outside the lab can read.

The founder knows the work is happening. But few outside the immediate team can really appreciate the scientific endeavour. There may be little tangible evidence to point to when an investor or even a board member asks what is currently being achieved.

Working at full intensity for months, without being able to demonstrate breakthrough progress externally, is a particular kind of strain. It wears founders down differently than ordinary startup pressure. The broader founder data picks this up. Sifted's 2025 survey of 138 founders found that 54% experienced burnout in the past 12 months and 75% reported anxiety. The pattern sharpens where timelines are longest: climate-tech founders described their mental health as "bad" or "very bad" at 63%, against 43% for everyone else.

The standard reading is that fundraising is hard and founders are exhausted. That reading misses the point. Burnout from too many investor meetings is one problem; burnout from running fast on a problem that refuses to tell you whether you are winning is a different problem entirely. The DeepTech founder sits disproportionately in the second category.

Resilience is not patience

The founder ecosystem talks a great deal about resilience. Resilience, properly defined, is the capacity to recover from setbacks: a churned customer, a failed hire, a lost round. It is real, it is necessary, and it is well-covered ground.

Patience is something else. Patience is the capacity to hold conviction during long stretches when nothing is happening at all, when there is no setback to recover from because there is no event. The two skills are easily confused because the surface symptom looks the same: a founder enduring. Underneath, the demands are different. Resilience requires processing a discrete loss and moving forward; the recovery has a beginning and an end. Patience requires sitting with sustained ambiguity, sometimes for years, without an external anchor and without a moment of resolution to push off from.

The danger inside this sustained ambiguity is that doubt becomes personal. The founder who confuses doubt about the technology with doubt about themselves collapses under it; the founder who treats doubt as engineering uncertainty stays in motion. Identity-as-founder and belief-in-the-bet need to be deliberately separated, so that questioning the technology does not feel like questioning the self. (The related engineer-to-CEO transition compounds this for technically-rooted founders.) Founders who cannot make this separation confuse the absence of signal with the absence of merit and quietly stop building. This is the failure mode that ends companies that should have survived.

The build-ship-learn loop fits a world of discrete signals and tight iteration; it fails in a world of sustained ambiguity. Telling a DeepTech founder to trust a process they cannot verify is asking them to substitute conviction for evidence. The two are not the same thing.

A Cambridge founder built his own evidence

Steve Brierley, CEO of Cambridge-based Riverlane, has spoken on record about what convinced him to start the company. In 2016, while a research fellow at Cambridge, Brierley sat at odds with the majority view in the global quantum community: that building a useful quantum computer "wouldn't be possible in our lifetime". The problem, his peers said, was simply too big and too hard.

Brierley disagreed, but he understood disagreement was not evidence. His response was operationally specific. As an outsider trying to map the state of the field, he approached experimental quantum research groups across other universities and the early commercial labs working on quantum hardware. He went team by team, asking each one how their capability that year compared with the year before, and the year before that. He was not looking for consensus; he was looking for a trendline.

What he found was a Moore's-Law-like doubling of capability every two years. He had no external validation, but he had built his own data, and that was enough to act on. He founded Riverlane the following year. The company has since closed Europe's first Series C in quantum computing, raising $75M to develop its quantum error correction stack. The instructive part is not the outcome; it is the discipline.

Building your own evidence

What Brierley did is not unique to quantum. It is the move strong DeepTech founders learn to make in any long-cycle technology. They stop waiting for the world to confirm the bet and start constructing evidence themselves.

In practice this means an internal scorecard, typically updated on a monthly cadence, anchored in milestones the founder defines and can credibly measure: a specific technical threshold, a specific customer conversation, a specific regulatory step. The scorecard is primarily for the founder. It is the mechanism through which they hold themselves accountable to their own standards, develop a clearer view of whether they are pursuing the right solution, and retain the ability to make independent decisions about whether to continue, pivot, or stop, even in the absence of investor scrutiny.

The beauty of this approach is that the scorecard does double duty. The same artefact that gives the founder internal confidence also becomes, in the right hands, a powerful external tool: a structural underpinning for board conversations and a credible exhibit at the next round.

In practice, the best DeepTech scorecards are designed with one eye on the milestones the founder needs to hit to credibly solicit investor interest at the right moments. The milestones the founder holds themselves accountable to become the milestones a serious investor will use to assess the journey. Building the scorecard for both audiences at once is what turns it from a private accountability mechanism into a fundraising asset.

When the investor drifts

If signal drought is the internal failure mode, then investor drift is the external one. The same Sifted survey found that 56% of founders received no mental health support from investors, and only 12% turned to investors when they needed it. Most experienced founders would say that is exactly as expected; serious founders rarely look to their investors for that kind of support in the first place. One founder put it plainly: "I've lost my faith in the stories from investors about supporting founders and come to realise they will do what they need to make money." The numbers confirm a reality the ecosystem already knows.

The harder problem is what they imply at long timescales. A typical VC fund deploys most of its capital in the first three to five years, after which partner attention shifts to new investments and the next vintage. For a DeepTech founder running an 8 to12-year mission, the investor whose conviction was bought in year one is statistically likely to be quietly disengaging by year six - if not before.

A second dynamic compounds the first. As later rounds come in, the early investors are diluted; new, larger shareholders take board seats and tend to set the agenda. The most engaged voices around the table in year six are rarely the ones who underwrote the original bet. This is not malice; it is fund mechanics and cap-table physics.

Funds are paid to deploy capital and return it, not to provide a decade of high-touch partnership. Founders who assume their lead investor will remain their primary commercial partner across the mission are pricing in a relationship the fund economics do not support.

Build the scaffolding yourself

If investors cannot reasonably provide a decade of intensive engagement, the DeepTech founder has to build it themselves. The discipline is to treat external support as commercial infrastructure, not wellbeing infrastructure, and to put it in place quietly and early, before it is needed.

In practice, four roles surface consistently as the ones that add real value. An executive coach, serving as a confidential sounding board on the CEO job itself. A founder mentor: a peer who has actually run a similar company, who can speak to lived experience rather than abstract frameworks. An industry advisor with genuine domain depth and customer credibility, not just a logo on the website. And an independent non-executive director who is not on the cap table and whose loyalty is to the company rather than any fund.

It is rarely all four at the same time, and the priority among them shifts with what the founder needs at each stage. Each fills a function that, structurally, an investor on a deployment cycle cannot. The cost feels optional when things are going well but the value is rarely disputed when things aren't.

Durable structures

The signal drought and the investor drift are really one problem in two dimensions, and they reinforce each other. A founder running low on conviction becomes more dependent on external validation, just as that validation is drifting away. A founder without scaffolding has nowhere to anchor conviction when the internal signal disappears.

The strongest DeepTech founders make two moves early. They build their own evidence, on their own scorecard, without investor prompting. And they build their own external scaffolding, deliberately and before they need it.

The ten-year founder and the five-year VC ecosystem will never run at the same speed. The founder who accepts that early and creates the structures to carry the missing years themselves, is the one still building when the technology finally talks back.

 
Let's talk.