The question for 2026 is no longer “should we do AI?” but “how do we move from PoC to industrialisation?”. 88% of AI PoCs never reach production — a wall that is 70% human, not technological. This white paper sets out the method: the seven frictions of the “last mile”, the five levers of change management, AgentOps, and a 12-month roadmap. It follows on from Maturity of AI in French SMEs 2025-2026.
Sourcing rule. Every claim is linked to a source. At most one third of the links point to internal content; at least two thirds point to external reference sources (Eurostat, BCG, McKinsey, MIT, RAND, IDC, HBR, the European Commission, and others).
1. June 2026: the year the question changed
For three years, the question SME leaders asked was: “Should we do AI?” That question is dead. Across Europe, Eurostat records 20.0% of enterprises using AI in the EU27 in 2025, up from 13.5% a year earlier. Everywhere, in every boardroom, the same question:
“We have a dozen PoCs. Some of them work. But nothing has scaled. How do we industrialise?”
This is the question of June 2026. According to IDC, 88% of AI PoCs never reach production; the RAND Corporation documents more than 80% failure before deployment; and Deloitte observes that 78% of companies are experimenting with agentic AI but only 13% have industrialised it.
The temptation is to read this as a technology gap. The BCG “10-20-70” rule disproves it: technology accounts for 10% of success, infrastructure and data for 20%, and people and processes for 70%. Moving from PoC to industrialisation is, by more than two thirds, a problem of change management — and, where agents are concerned, of AgentOps.
Go deeper (internal): The Junyr Method™ Scale: the five levels of AI maturity.
2. The European picture: 20% adoption, a skills wall
Eurostat puts enterprise AI adoption at 20.0% across the EU27 in 2025, with a dizzying gap by company size: 55% of large companies against 17% of small ones. The geographic divide is just as stark: according to rankings compiled from Eurostat data, Denmark leads at 42%, ahead of Finland (37.8%), Sweden (35%) and Belgium (34.5%), while Romania (5.2%), Poland (8.4%) and Bulgaria (8.5%) bring up the rear.
The real message comes down to a single figure: among the companies that considered AI but did not adopt it, Eurostat finds that 70.9% cite a lack of skills — ahead of regulatory uncertainty (52.5%) and data protection (48.8%). A Sharp survey of 2,500 SME leaders across ten European markets confirms it: 35% report teams anxious about their skills and 34% a distrust of AI’s outputs.
The technology is mature. What is missing is acculturation, trust and coordination.
Go deeper (internal): The end of prompt engineering.
3. The valley of death is not technical: the “last mile”
When a PoC fails to reach production, we look for the cause in the wrong place. That is the thesis of the landmark March 2026 article by Karim Lakhani, Jared Spataro and Jen Stave in Harvard Business Review — The “Last Mile” Problem Slowing AI Transformation: the obstacle “is rarely model quality or data availability, but the last mile where technical capability must meet organisational design.” Companies are “pilot-rich and transformation-poor.”
McKinsey finds that only 21% of organisations that have deployed generative AI have redesigned their workflows. And the MIT Project NANDA documents that more than 90% of employees already use personal AI tools (shadow AI), while 95% of organisations see no P&L impact.
To industrialise is to turn private, fragile, invisible use into collective, measured and governed use.
Go deeper (internal): The five fatal mistakes that doom 80% of AI transformations in SMEs.
4. The seven frictions of the last mile, seen from the field
The HBR article identifies seven structural frictions. Here is their on-the-ground translation for SMEs.
| ”Last mile” friction (HBR 2026) | On-the-ground reality in SMEs | Resolution lever |
|---|---|---|
| Pilot proliferation | A dozen PoCs, zero industrialisation | Tie every initiative to a business priority |
| Productivity gap | ”AI changes nothing in our numbers” | Adoption and impact KPIs before deployment |
| Process debt | Automating an already-broken process | ”Blank-page” redesign |
| Tribal-knowledge identity | The long-standing expert resists | Reposition them as an “architect” |
| Agentic governance | Agents with no control framework | AgentOps & a control plane |
| Architectural complexity | Stacked tools, scattered data | Multi-model orchestration, sovereignty |
| The efficiency trap | AI framed as cost-cutting | Reframe around value creation |
Beneath these frictions, I address five human resistances: fear of replacement, distrust of outputs (34% of European leaders), the skills deficit (70.9% of non-adopters), the absence of ownership and regulatory uncertainty.
Go deeper (internal): Junyr Agents™: delegating to AI without losing control.
5. The proof: what randomised controlled trials tell us
The argument that “change management creates value” is now backed by the highest level of evidence: the randomised controlled trial. A study in the Journal of Work-Applied Management — Impacts of adopting a new management practice: Operational Coaching™ — randomly assigned managers in English SMEs (40 trained SMEs versus 22 controls), under the ethical oversight of the London School of Economics and backed by the public Business Basics fund. The statistically significant result: training leaders to adopt a questioning, facilitative stance doubles the time they spend coaching their teams, with positive trends in productivity.
The SMEs’ long-standing productivity gap — a third lower than large firms across the OECD — is not closed by buying a tool, but by changing managerial posture.
6. The European framework as a lever — not a brake
The AI Act still applies from 2 August 2026 for transparency. But the political agreement on the Digital Omnibus has pushed back the heavier deadlines: according to Gibson Dunn, high-risk Annex III is deferred to 2 December 2027, and Annex I to 2 August 2028. Most SME use cases do not fall under high risk. Article 4 on AI literacy (in force since 2 February 2025) has been softened: to support rather than guarantee skills.
Compliance is an accessible floor — and the obligation to build AI literacy gives leaders a legitimate mandate to invest in training. The question “where does our data go?” turns the hosting choice into a lever for adoption.
Go deeper (internal): The “all-cloud” era is over · Cryptography 2026.
7. AgentOps: industrialising agents, not just deploying them
In 2026, we no longer deploy mere assistants but agents that perceive, reason and act. A MIT Sloan Management Review–BCG study finds that 76% of executives see agentic AI as a colleague. Yet you do not industrialise a synthetic colleague the way you industrialise a piece of software.
AgentOps is the discipline that extends DevOps/MLOps/LLMOps to observe and govern the lifecycle of autonomous agents in production. Red Hat describes it as a framework integrating observability, evaluation, governance, security and resilience. In concrete terms, as the 2026 agentic-observability tooling sets out, it is a control plane: real-time monitoring, policy enforcement, and guardrails able to halt a session or require human validation.
A concrete best practice: three LLM calls per step. Where a naïve agent chains its actions in a single call, Junyr Agents™ breaks each step into Plan → Execute → Verify/Test. This triptych, close to the “generator/critic” pattern, makes verification a native step: each action is gated by a test before the next, which curbs drift and makes the trajectory auditable. This is AgentOps made flesh — and the direct answer to distrust: you trust an agent that proofreads itself. The cost (three calls) is a FinOps trade-off managed through model routing.
Go deeper (internal): Junyr Agents™: delegating to AI without losing control.
8. Change management in five levers (the Junyr Method™)
The move from Artisan to Orchestra is the move from PoC to industrialisation. Five levers:
- Mandate and narrative — the leader carries the meaning (the “leader-as-teacher” posture).
- Business champions — one credible champion per process creates ownership.
- Three-tier training (ch. 9), the antidote to the skills wall.
- Minimum viable governance + AgentOps — a charter, a data policy, a control plane.
- Adoption measurement (ch. 10).
The golden rule: no level is skipped. An Artisan-level SME that sinks €200,000 into an agent platform fails nine times out of ten, for want of foundations.
Go deeper (internal): The Junyr Method™ Scale.
9. The three-tier training plan & the augmented leader
Tier 1 — Acculturation (everyone): dispel fear, set the frame. A building block required by Article 4 of the AI Act, and the answer to the 35% of anxious teams.
Tier 2 — Practitioners: to command AI rather than talk to it.
Tier 3 — Architects: to decide and to govern (use cases, ROI, AgentOps, compliance).
To this is added the work on the leader: turning the cognitive overload tied to burnout into an “augmented leader.” Drawing on France Num and Bpifrance’s “Osez l’IA” (Dare AI) plan (€10 billion).
Go deeper (internal): The end of prompt engineering.
10. Measuring adoption: KPIs and the agentic control plane
Three families of indicators, defined before deployment: adoption (weekly active users); impact (time saved, ROI — a median of ~159% over 12 months, whereas MIT NANDA shows 95% obtain none); and control (FinOps & AgentOps: inference cost, human validation, compliance score).
Go deeper (internal): Token budgets and AI APIs: the FinOps guide for SMEs in 2026.
11. The deflation of expertise: why value is shifting
Karim Lakhani’s work (HBS) puts it plainly: AI lowers the marginal cost of expertise. What once made a specialist SME scarce becomes abundant. The right response is not to endure it but to redeploy towards what AI does not commoditise: judgement, empathy, non-standard problems, storytelling, trust.
Go deeper (internal): Where is AI’s value heading? Three signals of commoditisation.
12. INDUSTEC, from the inside: the human mechanics of a 182% ROI
INDUSTEC — an industrial SME with 78 staff — posts 182% ROI in nine months, 275 hours saved per month, +18% revenue, and 1.7 FTE redeployed. What made the difference was not the technology: a narrative carried by the leader, three business champions trained before deployment, adoption indicators tracked every week, and light but genuine governance. Nine months on, the teams were using the tool every day, with confidence.
Go deeper (internal): INDUSTEC: 182% ROI in nine months.
13. A 12-month roadmap & European funding
Reference budget (SMEs and mid-caps of 50-500 staff): €30,000 to €80,000 over 12 months.
- Months 0-1 — Framing and narrative.
- Months 2-3 — Human foundations (champions, acculturation, charter; data audit €3,000-8,000).
- Months 4-7 — Industrialising the first use case (“blank-page” redesign, the 21% who make the difference).
- Months 8-10 — Extension and orchestration (agents under supervision with AgentOps).
- Months 11-12 — Measurement and consolidation.
Funding in Europe: Bpifrance Diag Data IA and IA Booster (25-50%), and the European Digital Innovation Hubs (Digital Europe programme).
14. Conclusion: to industrialise is to transmit
The wall between experimentation and everyday use — the one on which 80% of projects and 88% of PoCs founder — is not made of silicon. Technology accounts for 10%. The rest is change-management work, now validated by randomised controlled trials. To industrialise AI is to transmit a capability to an entire organisation — humans and agents alike.
Take action
The AI Express Diagnostic — 60 minutes, no commitment. Book a diagnostic · The first white paper · The blog.
Paul-Antoine TUAL — AI Transformation Leader · Engagements delivered through the SAS Croissance et Transitions.