The CEO Is Not VC’s Control Group
How should venture capital assess deep tech risk when science, scale-up and execution follow different timelines?
This article is a result of countless conversations inside of the field. The leading argument was inspired and co-guided by Kent Langel from Lanmer Group.*
Venture capital helped build transformative technology companies. But that history does not mean every modern VC process can distinguish scientific uncertainty from management failure.
In deep tech, a missed milestone can have several causes: the underlying phenomenon may be weaker than expected; the engineering architecture may be wrong; scale may introduce new constraints; production yield may remain too low; customers may resist adoption; or the team may simply execute badly. These outcomes can look identical in a board report while demanding completely different responses.
The governance problem begins when leadership becomes the easiest variable to change. A board cannot replace thermodynamics, negotiate with biological variation or vote away material degradation. It can replace the CEO. That makes decisive action visible, but visibility is not the same as diagnostic accuracy.
Why does conventional VC governance misread deep tech risk?
Institutional venture capital works best when uncertainty can be divided into affordable stages, tested quickly and translated into evidence before the next financing decision. Software became unusually compatible with this logic: experiments are comparatively cheap, products can change rapidly and user behaviour generates continuous information.
Physical and science-based ventures often operate differently. Resolving one uncertainty may require specialist facilities, scarce materials, regulatory interaction, customer integration or repeated production. A technically successful prototype may still reveal little about manufacturability, reliability, certification or industrial economics.
This is why VC due diligence for deep tech industrialisation must go beyond a market narrative and a technical expert call. The evidence-production system itself is expensive, and the next meaningful result may arrive after eighteen months rather than eighteen days.
Why is product-market fit insufficient for deep tech investment readiness?
Customer demand still matters. Scientific elegance cannot rescue a technology nobody needs. The error is treating product-market fit as a universal measure of progress.
A deep tech company may need several kinds of fit at the same time:
Phenomenon–application fit: Does the scientific effect solve a valuable problem?
Technology–system fit: Can it operate inside the customer’s technical environment?
Process–manufacturing fit: Can it be produced reproducibly and at the required yield?
Cost–adoption fit: Can customers justify qualification, switching and operating costs?
Capital–timeline fit: Can the decisive evidence be financed in the correct sequence?
A company can have enthusiastic pilot customers and still fail at scale. It can have a working device and fail certification. It can meet a technical specification yet prove impossible to integrate into an operating plant. Scale is not always repetition; scale can change the behaviour being repeated.
Why do funding rounds fail as deep tech traction metrics?
Financing rounds are highly visible. They have dates, amounts, investors and valuations. Production qualification, regulatory clearance and durable customer adoption are harder to compare and disclosed unevenly. As a result, another investment can become the default proxy for progress even when the decisive transition lies elsewhere.
That creates circular evidence. Investors use earlier investment as a signal. Public programmes report follow-on capital as impact. Companies optimise milestones for the next investor. A successful round then appears to validate the company, the fund and the support programme.
Capital may preserve option value and fund the next experiment, but it is not proof that production, regulation, adoption or unit economics have advanced. Our research on venture capital continuity and hard commercialisation separates financing continuity from strict commercial outcomes for exactly this reason.
When is replacing a deep tech CEO the right decision?
Leadership changes can be necessary. Founders can allocate capital badly, conceal negative results, hire poorly, pursue the wrong market or refuse to abandon a falsified thesis. Accountability is essential.
The decision should follow diagnosis, not substitute for it. A deep tech board should
ask what uncertainty the milestone was designed to resolve, whether the experiment failed or exposed a different mechanism, which constraint sits on the critical path and whether more capital can genuinely accelerate it. It should define in advance which evidence would justify changing the product, the financing architecture, the leadership or the company itself.
Without those distinctions, governance converts scientific ambiguity into personnel decisions. The CEO becomes the control group because nature refuses to be one.
What should deep tech fund due diligence measure instead?
Deep-tech-native due diligence separates scientific, engineering, scale-up, economic, market and execution risk. It treats pilots as evidence rather than proof of scalability and measures learning, not only positive results. A failed experiment that eliminates an entire technical route may create more decision value than a polished demonstration that leaves the binding uncertainty untouched.
It also gives specialist knowledge authority. An adviser whose judgment can be ignored by a generalist investment committee does not create governance capability. The decision system needs explicit evidence thresholds, owners and escalation paths.
Finally, the capital instrument should match the uncertainty. Grants may fund scientific validation. Customer-funded development can establish application relevance.
Strategic partners may provide manufacturing and integration capability. Equipment finance or project capital may fund assets that do not belong on a venture balance sheet. VC becomes powerful when capital can accelerate a sufficiently repeatable technical and commercial path.
How can investors build a better deep tech decision system?
There is no universal funding architecture. The practical task is to identify the next value-changing uncertainty, define the evidence required to resolve it and choose governance and capital that can hold the real timeline.
At Arise Innovations, I work with investors, institutions and science-based ventures to distinguish technology risk from execution risk, test whether milestones reflect industrial reality and design evidence frameworks around the specific asset. Our research and intelligence work supports decisions where generic startup metrics are too coarse for the technology being assessed.
Deep tech does not need investors who never lose conviction. It needs investors whose conviction can be updated by the right evidence. The CEO remains accountable for the company’s response—but the CEO is not the control group, and nature is not an employee. Explore the full argument:
*No commission or commercial interest is represented here.
Kommentare