What Separates Organizations That Can Actually Absorb Deep Tech?

Many organizations now claim to care about deep tech. They reference AI in strategic decks, attend ecosystem events, launch innovation initiatives, and run occasional pilots. But far fewer have built the institutional machinery required to identify, evaluate, fund, pilot, deploy, and learn from frontier technologies in a repeatable way. That is the real dividing line: not between interest and disinterest, but between visible activity and institutional absorption.

This distinction matters because deep tech does not move through organizations the way lightweight digital tools often do. It is more capital-intensive, more uncertain, more technically demanding, and usually harder to integrate into existing operations. In that context, absorption is not a rhetorical posture. It is an organizational capability: one that determines whether a promising technology becomes operational reality, whether a pilot converts into scaled deployment, and whether the organization can learn its way through long-horizon uncertainty rather than performing innovation for symbolic effect. The most useful framing is this:

That is what separates a genuine absorber from an organization trapped in pilot theater. A practical way to frame the problem is as a minimum viable absorptive system: the minimum set of institutional conditions required for deep tech engagement to become real.

The concept of absorptive capacity, first articulated by Cohen and Levinthal in 1990, defines it as the ability to recognize the value of new external information, assimilate it, and apply it to commercial ends. Subsequent work by Zahra and George (2002) refined this into two components:

• Potential Capacity: the ability to scout and recognize promising technology.
• Realized Capacity: the ability to transform and exploit it.

Organizations performing innovation theater typically have the first. They lack the second. Evidence suggests that nearly 85% of AI initiatives never make it to production, and 95% of generative AI pilots fail to deliver measurable financial returns, rarely because of technological insufficiency.

The Six Conditions That Separate Absorbers from Performers

1. Strategic Anchoring

Organizations that absorb deep tech do not treat it as a side project owned by a single innovation team. They anchor it in strategy, executive sponsorship, and governance. Research by Chatterjee et al. (2021) consistently shows that leadership support is a necessary condition for moving beyond pilots, and that fragmented ownership is a recurrent cause of stalled initiatives. In practice, strength looks like formal sponsorship, clear decision rights, and review mechanisms that connect deep tech activity to real priorities. Weakness looks like ambition without accountability: deep tech appears in the language of the organization, but not in the operating commitments that govern resource allocation and follow-through. Only 28% of organizations currently have CEO-level AI governance, yet this is the strongest predictor of long-term returns.

2. Resourcing Seriousness

Deep tech absorption requires more than episodic project funding. Dedicated budgets, structured financing mechanisms, and the ability to sustain activity over time are essential conditions, as Agostini and Nosella (2019) and subsequent empirical work confirm. The underlying issue is not simply whether money exists, but whether the organization has built a credible capital pathway from early exploration to pilot and from pilot to scale. Meaningful absorbers replace storytelling-driven funding with milestone-based capital sequencing: catalytic capital in the first year to de-risk the science, pilot funding in years one to two to prove repeatability, and scale-up capital beyond that to fund the transition to operating conditions. Weak organizations authorize experimentation in principle but starve it in practice. This is one reason pilot-heavy organizations can still be structurally immature.

3. Technical Validation Capability

Deep tech is not absorbed by enthusiasm alone. It requires internal ability to judge technical credibility, assess fit, and manage the transition from proof to production. Roberts (2015) finds that data integration and technical skill depth matter more than surface-level partnerships. Many firms can acquire access to a technology; only those with robust transformation routines can bridge the gap between a lab-scale prototype and a production-grade asset. In pharmaceuticals, for instance, the Manufacturing Science and Technology (MSAT) function serves as precisely this institutional validation bridge, ensuring new processes are robust, reproducible, and regulatory-compliant before they reach manufacturing. Organizations that lack an equivalent validation function leave promising technology stranded upstream of production.

4. Execution Adaptability, Especially Around Procurement

One of the clearest findings across the research is that pilot-to-scale failure is often less about technical insufficiency than about organizational drag. Slow decision cycles, rigid procurement, and unclear progression criteria create what the literature calls pilot purgatory: projects that succeed technically but never convert into contracts. McKinsey and Gartner estimate that between 70% and 90% of projects fail to scale beyond the pilot phase. An organization may have strategy, budget, and partnerships; but if it cannot contract quickly enough, define clear go or no-go gates, or create fast-track pathways for non-traditional suppliers, it cannot absorb deep tech at operating speed. Procurement is not a back-office function. It is part of the absorption architecture. The Venture Client model addresses this directly: as documented in the State of Venture Clienting 2024, successful Venture Client Units issue purchase orders in under 12 weeks, and 71% of them report within two levels of the C-suite to ensure the necessary resources are available when a pilot reaches a critical gate.

5. Business-Unit Absorption and Commercialization Logic

Deep tech maturity is not proven when an innovation team completes a pilot. It is proven when the rest of the institution can own the result. Kowal and Przewoźnik (2025), examining deep tech ecosystems across Europe, confirm that weak business-unit adoption is a recurrent bottleneck, especially in large organizations. Deep tech failure often appears after technical feasibility has already been demonstrated: the problem is not whether the technology works in isolation, but whether it can be integrated into core systems, aligned with operating incentives, and linked to a credible value logic. Absorption is incomplete until technology can survive the institution's immune system and become part of how the business actually functions. In Europe, this is compounded by the fact that many technically strong ventures underperform on the corporate customer side not because the science is weak, but because pilot-to-contract conversion and commercialization discipline are weak.

6. Learning Capture

Mature organizations do not simply run experiments. They retain and reuse what those experiments reveal. Alexiou, Khanagha, and Schippers (2019) identify continuous learning mechanisms and cross-functional knowledge transfer as important differentiators of absorptive capacity. The deeper problem is that failed pilots are often poorly instrumented: when organizations cannot explain why an initiative failed, the failure is paid for twice, once in wasted effort, and again in lost institutional learning. Strength here looks like observability, structured feedback loops, and mechanisms that push insights from one initiative into future decisions. Weakness looks like isolated project memory, informal handover, and repeated mistakes under new branding. In deep tech, learning is not an auxiliary benefit. It is part of the operating model.

Where Breakdowns Actually Happen

Taken together, these conditions reveal where breakdowns usually occur. The decisive failures are often not within a function but between functions: strategy not translated into budget, scouting not matched by technical validation, pilot success not producing procurement movement, innovation teams unable to secure business-unit ownership, and experiments completed but lessons not institutionalized. This is why deep tech absorption is inherently cross-functional, and why visible activity can be so misleading. A company can look busy, connected, and ambitious while remaining structurally unable to move deep tech through the handoffs that determine whether it sticks.

This is also why averages conceal the truth. Two organizations may both appear active in deep tech: both mention frontier technologies in strategy documents, both collaborate externally, and both carry a portfolio of pilots. Yet one may have aligned sponsorship, disciplined validation, adaptive procurement, business-unit pull, and learning loops, while the other has fragmented ownership, opportunistic resourcing, slow contracting, weak technical judgment, and no scale-up pathway. At headline level they look similar. Structurally they are different institutions. One is building capability. The other is performing readiness.

The European Dimension

The European context makes this distinction especially consequential. Europe is a research powerhouse with world-class universities and engineering institutions, yet it consistently lags in the adoption and commercialization of its own technologies. The EIB Investment Report 2024/2025 identifies a largely bank-based financial system, persistent market fragmentation, and a scale-up financing gap as structural barriers, with technologies invented in Europe often industrialized elsewhere. Approximately 60% of exporting European firms cite intra-EU market fragmentation as a primary obstacle to competitiveness. Where the external environment already makes scaling harder, internal institutional weakness becomes even more expensive. In that setting, rhetoric is especially cheap and operating mechanisms especially decisive.

The Better Question

The relevant executive question is not: "Are we active in deep tech?" Nor is it: "Do we have a strategy slide on AI, biotech, advanced materials, or energy transition?" The better question is harder and more useful: do we have the institutional conditions required to absorb frontier technologies repeatedly and turn them into operating capability or commercial value?

If the answer is unclear, visible activity should not be taken as reassurance. In deep tech, readiness is structural. And structural weakness has a way of revealing itself precisely when the stakes become real.

---

Bibliography

1. Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128–152. View source
2. Zahra, S. A., & George, G. (2002). Absorptive capacity: A review, reconceptualization, and extension. Academy of Management Review, 27(2), 185–203.
3. Chatterjee, S., Rana, N., Dwivedi, Y., & Baabdullah, A. (2021). Understanding AI adoption in manufacturing and production firms using an integrated TAM-TOE model. Technological Forecasting and Social Change, 170, 120880. https://doi.org/10.1016/j.techfore.2021.120880
4. Agostini, L., & Nosella, A. (2019). The adoption of Industry 4.0 technologies in SMEs: Results of an international study. Management Decision. https://doi.org/10.1108/md-09-2018-0973
5. Roberts, N. (2015). Absorptive capacity, organizational antecedents, and environmental dynamism. Journal of Business Research, 68, 2426–2433. https://doi.org/10.1016/j.jbusres.2015.02.019
6. Alexiou, A., Khanagha, S., & Schippers, M. (2019). Productive organizational energy mediates the impact of organizational structure on absorptive capacity. Long Range Planning. https://doi.org/10.1016/j.lrp.2018.02.001
7. Kowal, D., & Przewoźnik, W. (2025). Deep tech ecosystems as drivers of sustainable development: Entrepreneurship and innovation perspectives from Europe and Poland. Sustainability. https://doi.org/10.3390/su172210195
8. 27pilots. (2024). State of Venture Clienting 2024. stateofventureclient.com
9. European Innovation Council. (2025). EIC Impact Report 2025: Scaling Deep Tech in Europe. View report
10. European Investment Bank. (2024). EIB Investment Report 2024/2025: Innovation, Integration and Resilience. View report
11. InformationWeek. (2025). From pilot purgatory to productive failure: Fixing AI's broken learning loop. View article
12. The Data Experts. (2025). The enterprise AI crisis: Why 95% of AI projects fail and how to join the 5% that succeed. View article
13. Economic Times Manufacturing. (2025). Why deep tech can't move at software speed. View article
14. OXB Careers. Manufacturing Science and Technology (MSAT) and Validation. View source
15. EdTech Digest. (2025). From pilot to scale: Why most AI projects fail to move the needle. View article