top of page
Research & intelligence
Research for decisions that cannot wait for certainty.
Original research and analysis on how scientific evidence, institutional incentives and capital shape commercialisation.
Explore programme evaluation, technology adoption and the decisions that connect research to deployment.
Research, Frameworks & Industry Reports
*For full description & link to download - click on the item.
4×4-TETRA Deep Tech Matrix
Core insight: A strong scientific result becomes a viable venture only when evidence, industrialisation, market development and capital are sequenced together.
The challenge: Progress in one area can create constraints in another; technology readiness alone does not capture these interactions.
The approach: TETRA maps the venture’s changing state across time and technology readiness levels.
The planning logic: It works backwards from a target outcome to identify the evidence and decisions needed along the way.
The implication: Founders, investors and institutions can examine structural misalignment before an early financing or scaling decision closes off options.
The challenge: Progress in one area can create constraints in another; technology readiness alone does not capture these interactions.
The approach: TETRA maps the venture’s changing state across time and technology readiness levels.
The planning logic: It works backwards from a target outcome to identify the evidence and decisions needed along the way.
The implication: Founders, investors and institutions can examine structural misalignment before an early financing or scaling decision closes off options.
Structural Interaction of KPI Architectures
Core insight: A programme’s KPIs help shape its outcomes long before those outcomes are measured.
The challenge: Short reporting cycles can reward visible activity while scientific and industrial validation take longer.
The approach: The paper models interactions among KPI weightings, reporting cycles and institutional mandates using four domains and 16 indicators.
The finding: Its simulation illustrates how misaligned measures can produce path-dependent distortions across programme cycles.
The implication: Align KPIs with deep tech development before a mandate begins, when the programme can still change its trajectory.
The challenge: Short reporting cycles can reward visible activity while scientific and industrial validation take longer.
The approach: The paper models interactions among KPI weightings, reporting cycles and institutional mandates using four domains and 16 indicators.
The finding: Its simulation illustrates how misaligned measures can produce path-dependent distortions across programme cycles.
The implication: Align KPIs with deep tech development before a mandate begins, when the programme can still change its trajectory.
Deep Tech Playbook
Core insight: In deep tech, the order of decisions can matter as much as the decisions themselves.
The system: Scientific evidence, industrialisation, market development and capital continually influence one another.
The risk: Premature financing, manufacturing commitments or market promises can lock a venture into a difficult path.
The approach: The playbook provides a map for examining dependencies, trajectories and decision timing.
The implication: Founders, investors and institutions can connect funding and governance decisions to the venture’s actual technical milestones.
The system: Scientific evidence, industrialisation, market development and capital continually influence one another.
The risk: Premature financing, manufacturing commitments or market promises can lock a venture into a difficult path.
The approach: The playbook provides a map for examining dependencies, trajectories and decision timing.
The implication: Founders, investors and institutions can connect funding and governance decisions to the venture’s actual technical milestones.
Beyond Venture Capital
Core insight: The right capital depends on the venture’s current risks and the evidence it needs to produce next.
The evidence base: The report examines 100 documented companies and identifies ten recurring capital architectures beyond formative institutional venture capital.
The pattern: Revenue, procurement, grants, licensing and strategic partnerships can finance development while testing real demand or technical performance.
The boundary: Venture capital becomes a stronger fit when rapid deployment, repeatable sales and a credible liquidity path are plausible within a fund’s time horizon.
The implication: Choose financing for the next scarce capability, while preserving valuable future options.
The evidence base: The report examines 100 documented companies and identifies ten recurring capital architectures beyond formative institutional venture capital.
The pattern: Revenue, procurement, grants, licensing and strategic partnerships can finance development while testing real demand or technical performance.
The boundary: Venture capital becomes a stronger fit when rapid deployment, repeatable sales and a credible liquidity path are plausible within a fund’s time horizon.
The implication: Choose financing for the next scarce capability, while preserving valuable future options.
Do Startup Programmes Actually Commercialise Science?
Core insight: Startup formation and fundraising should not be mistaken for demonstrated science commercialisation.
The measurement gap: Programme evaluations often track early signals more consistently than industrial deployment or sustained commercial outcomes.
The pilot observation: In a sample of 20 European companies meeting defined commercialisation thresholds, accelerator participation is publicly documented for two.
The caution: The sample is exploratory, and a missing public record does not establish that a company never participated in a programme.
The implication: Assess programmes against long-term commercialisation outcomes and examine the roles of spinouts, industrial partners, licensing and revenue.
The measurement gap: Programme evaluations often track early signals more consistently than industrial deployment or sustained commercial outcomes.
The pilot observation: In a sample of 20 European companies meeting defined commercialisation thresholds, accelerator participation is publicly documented for two.
The caution: The sample is exploratory, and a missing public record does not establish that a company never participated in a programme.
The implication: Assess programmes against long-term commercialisation outcomes and examine the roles of spinouts, industrial partners, licensing and revenue.
Generic vs. Specialised Programmes
Core insight: A programme creates value when it addresses the venture’s dominant uncertainty.
The mismatch: Generic accelerators often focus on market discovery, while science ventures at TRL 3–6 may be constrained by technical validation, scale-up, regulation or IP.
The mechanisms: A poor fit can divert development time, raise the cost of pivots, increase disclosure risk and narrow future application options.
The evidence limit: The paper’s cost estimates are illustrative scenarios; it does not establish a causal advantage for specialised programmes.
The implication: Identify each venture’s binding constraint at intake, then match support and evaluation to that constraint.
The mismatch: Generic accelerators often focus on market discovery, while science ventures at TRL 3–6 may be constrained by technical validation, scale-up, regulation or IP.
The mechanisms: A poor fit can divert development time, raise the cost of pivots, increase disclosure risk and narrow future application options.
The evidence limit: The paper’s cost estimates are illustrative scenarios; it does not establish a causal advantage for specialised programmes.
The implication: Identify each venture’s binding constraint at intake, then match support and evaluation to that constraint.
From Description to Decision in Deep Tech
Core insight: A deep tech framework becomes useful for governance when it helps people decide what to do.
The gap: Lists of attributes improve description but may leave thresholds, borderline cases and measurement unresolved.
The tool: The paper introduces the Decision-Capability Test, covering thresholds, boundaries, operationalisation, dynamics and governance.
The finding: Descriptive richness does not automatically translate into the ability to classify ventures or guide institutional action.
The implication: A deep tech classification should inform funding rules, programme design, KPIs and oversight decisions.
The gap: Lists of attributes improve description but may leave thresholds, borderline cases and measurement unresolved.
The tool: The paper introduces the Decision-Capability Test, covering thresholds, boundaries, operationalisation, dynamics and governance.
The finding: Descriptive richness does not automatically translate into the ability to classify ventures or guide institutional action.
The implication: A deep tech classification should inform funding rules, programme design, KPIs and oversight decisions.
Access Before Ownership
Core insight: Deep tech’s early capital needs depend partly on whether ventures can access essential resources before they must own them.
The distinction: A technology may require expensive facilities or specialist expertise without requiring the startup to put every asset on its own balance sheet.
The European challenge: Research infrastructure exists, but fragmented access, contracting delays and IP practices can make it difficult for young companies to use.
The boundary: Shared access cannot replace ownership in every case or guarantee a lower total cost.
The implication: Make facilities, specialist skills and services easier to find, contract and use. This is a working hypothesis that calls for comparative testing.
The distinction: A technology may require expensive facilities or specialist expertise without requiring the startup to put every asset on its own balance sheet.
The European challenge: Research infrastructure exists, but fragmented access, contracting delays and IP practices can make it difficult for young companies to use.
The boundary: Shared access cannot replace ownership in every case or guarantee a lower total cost.
The implication: Make facilities, specialist skills and services easier to find, contract and use. This is a working hypothesis that calls for comparative testing.
The Asset Changed Faster Than the Institution
Core insight: Venture funds may move into hardware deep tech faster than they develop the ability to select, finance and govern it well.
The shift: Generative AI can shorten early software experimentation, while physical validation, manufacturing and certification still take time.
The tension: Hardware ventures can absorb substantial capital, but that does not make venture capital the right instrument at every stage.
The hypothesis: A gap between a fund’s investment thesis and its capabilities may favour ventures that look familiar to software investors over those with stronger technical-commercial pathways.
The implication: Evaluate an investor’s decision process and capital model alongside its stated deep tech thesis. The paper develops a testable theory, not a new causal finding.
The shift: Generative AI can shorten early software experimentation, while physical validation, manufacturing and certification still take time.
The tension: Hardware ventures can absorb substantial capital, but that does not make venture capital the right instrument at every stage.
The hypothesis: A gap between a fund’s investment thesis and its capabilities may favour ventures that look familiar to software investors over those with stronger technical-commercial pathways.
The implication: Evaluate an investor’s decision process and capital model alongside its stated deep tech thesis. The paper develops a testable theory, not a new causal finding.
Venture Capital Continuity and Hard Commercialisation
Core insight: A venture’s ability to raise another round and its progress toward hard commercialisation are different outcomes.
The study: This exploratory analysis follows 48 European physical-technology ventures selected by the EIC Accelerator in 2021.
The financing pattern: A qualifying follow-on round was observed for 63.6% of ventures with confirmed prior VC, compared with 34.6% without confirmed prior VC.
The commercialisation pattern: Among ventures that were pre-commercial at the starting point, the association between prior VC and reaching the study’s hard commercial benchmark was smaller and statistically imprecise.
The implication: Investors and policymakers should measure financing continuity and commercial progress separately. The observational study does not establish a causal VC effect.
The study: This exploratory analysis follows 48 European physical-technology ventures selected by the EIC Accelerator in 2021.
The financing pattern: A qualifying follow-on round was observed for 63.6% of ventures with confirmed prior VC, compared with 34.6% without confirmed prior VC.
The commercialisation pattern: Among ventures that were pre-commercial at the starting point, the association between prior VC and reaching the study’s hard commercial benchmark was smaller and statistically imprecise.
The implication: Investors and policymakers should measure financing continuity and commercial progress separately. The observational study does not establish a causal VC effect.
Advanced Materials Act: Stakeholder Feedback
Core insight: Europe’s advanced materials policy should be judged by repeatable industrial adoption, not only by research activity or the number of programmes launched.
The gap: Promising materials often struggle to move from laboratory results through qualification and pilot production to paying customers.
The proposed route: Shared pilot infrastructure, faster qualification and recognised test data can make adoption more practical.
The demand side: SME-friendly contracts, paid pilots, procurement and risk-sharing can help industrial customers become early adopters.
The implication: Track outcomes such as time to first industrial deployment and the share of pilots that become recurring commercial orders.
The gap: Promising materials often struggle to move from laboratory results through qualification and pilot production to paying customers.
The proposed route: Shared pilot infrastructure, faster qualification and recognised test data can make adoption more practical.
The demand side: SME-friendly contracts, paid pilots, procurement and risk-sharing can help industrial customers become early adopters.
The implication: Track outcomes such as time to first industrial deployment and the share of pilots that become recurring commercial orders.
Laundering the Quantum Horizon
Core insight: A quantum market forecast is difficult to assess when its technology boundary, metric or timeline changes without a clear explanation.
The audit: The authors compare seven public reports from two advisory report families published between 2022 and 2026.
The measurement problem: Provider revenue, end-user economic value, investment, valuation and public funding describe different things; combining them can blur commercial maturity.
The evidence boundary: The analysis tests what readers can trace in public documents. It does not assess proprietary models or the intentions of report authors.
The implication: Forecasts should separate technology domains and metrics, state their evidence clearly and reconcile changes between editions.
The audit: The authors compare seven public reports from two advisory report families published between 2022 and 2026.
The measurement problem: Provider revenue, end-user economic value, investment, valuation and public funding describe different things; combining them can blur commercial maturity.
The evidence boundary: The analysis tests what readers can trace in public documents. It does not assess proprietary models or the intentions of report authors.
The implication: Forecasts should separate technology domains and metrics, state their evidence clearly and reconcile changes between editions.
Fusion, Deep Tech, and Who Owns Which Handover?
Core insight: Fusion commercialisation depends on whether ventures, suppliers, infrastructure providers and customers can make their capabilities and commitments work together.
The handover problem: Each transfer carries specifications, evidence, operating assumptions and responsibility—not just technology.
The coordination challenge: Funding, manufacturing capacity, a site, regulatory progress, grid access and customer demand can advance on different timelines. Progress in one does not complete the others.
The wider pattern: Fusion cases and examples from semiconductors, industrial biotechnology and quantum hardware illustrate how technical capability depends on its surrounding industrial system.
The implication: Ask who owns each handover and where the ability to repeat the work accumulates. The report uses the cases to examine interdependence, not to predict company success.
The handover problem: Each transfer carries specifications, evidence, operating assumptions and responsibility—not just technology.
The coordination challenge: Funding, manufacturing capacity, a site, regulatory progress, grid access and customer demand can advance on different timelines. Progress in one does not complete the others.
The wider pattern: Fusion cases and examples from semiconductors, industrial biotechnology and quantum hardware illustrate how technical capability depends on its surrounding industrial system.
The implication: Ask who owns each handover and where the ability to repeat the work accumulates. The report uses the cases to examine interdependence, not to predict company success.
Newsletter
Briefs
Analysis of the decisions connecting science, commercialisation, capital and policy. Each brief examines a consequential question, the evidence behind it and what it could change about the next move.
Deep Dives, Not Hot Takes
Who Taught the People Teaching Founders?
Deep tech startup advice is only useful when experience matches the decision. An exploratory review of twelve organisations asks how founders can identify relevant expertise—and whether that expertise can challenge the programme’s chosen path.
Maria Ksenia Witte
Programme Additionality: When Busy Innovation Programmes Change Nothing
Programme additionality asks what your support changes—not how much activity it produces. A practical perspective for institutions deciding what to renew, redesign or stop.
Maria Ksenia Witte
Innovation Programme Evaluation: Is Your Institution Part of the Problem?
Strong activity does not automatically justify another funding cycle. Why innovation programme evaluation must examine the institution’s own behaviour, credibility and delivery conditions before leaders renew, redesign or stop support.
Maria Ksenia Witte
The Gravity of Default Architectures
Institutions often respond to novel technologies with familiar structures. This analysis explains how default architectures, path dependence and evidence filters shape deep tech commercialization and institutional change.
Maria Ksenia Witte
Why Deep Tech Commercialisation Has a Free Work Problem
Free expert labour is not just a budget choice. It shapes who gets heard, what founders learn about value, and whether publicly funded commercialisation programmes create real economic accountability.
Maria Ksenia Witte
Why Strong Deep Tech Founders Avoid Accelerator Programmes—and How to Fix the Signal
When experienced deep tech founders avoid a programme, their absence is a signal. They look for industrial outcomes, operational access and evidence of commercialization.
Maria Ksenia Witte
Europe’s Innovation Gap Is a Handover Problem
Europe has research, capital and policy instruments. The innovation gap increasingly appears in the handovers between them—where evidence, ownership and momentum are lost.
Maria Ksenia Witte
Why Deep Tech Support Systems Optimize for Continuation, Not Conversion
Many deep tech programmes are excellent at producing cohorts, workshops and demo days. The harder test is whether they convert scientific assets into industrial deployment.
Maria Ksenia Witte
Why Deep Tech Accelerator Programs Fail — and What to Build Instead
Deep tech accelerators often apply software startup logic to scientific uncertainty. Better programmes align technology validation, regulation, manufacturing, capital and deployment.
Maria Ksenia Witte
Why Institutional Power Without Accountability Damages Innovation
When authority is separated from consequences, institutions can punish operational competence and lose the feedback needed for better technology and innovation decisions.
Maria Ksenia Witte
bottom of page