Research note / 28 September 2026
Ten public observations shaped the Decision Sprint. Each item separates published evidence from our interpretation. The sources describe different populations and should not be combined into one market estimate.
Evidence. The OECD's 2025 D4SME survey reports maintenance costs as a digitalisation barrier for 40% of respondents, ahead of hardware costs at 32%.
Interpretation. A useful engagement must make operating cost and ownership visible before recommending a system.
Evidence. The same survey records lack of time for training as a barrier for 39% of SMEs and 49% of medium-sized businesses.
Interpretation. The first step should be asynchronous and bounded, with a short evidence request rather than open-ended workshops.
Evidence. In the OECD sample, 35% used internet research and 23% colleague knowledge-sharing; 13% used formal training and 6% expert-led programmes.
Interpretation. Delivery should leave a reusable decision record and operating checklist, not depend on a presentation or hidden expertise.
Evidence. Only 4% of respondents said they seek external experts to organise or implement their digitalisation process.
Interpretation. A small, reversible and inspectable first engagement is more credible than a broad transformation promise.
Evidence. The OECD survey found no clear digitalisation process at 45% of businesses led by CEOs aged 65 or older, versus 11% for leaders aged 25–34. This is an association in the sample, not proof that age causes the difference.
Interpretation. The method should work independently of management style by making options, authority and review gates explicit.
Evidence. AI use in the OECD SME sample rose from 26% in 2024 to 39% in 2025. Non-users cited legal issues, privacy and misinformation among deterrents.
Interpretation. Recommendations need source provenance, explicit uncertainty, data-minimisation and a named human approval point.
Evidence. Eurostat reported that 13.5% of EU enterprises with at least ten workers used AI in 2024, up from 8.0% in 2023. This population differs from the OECD SME sample above.
Interpretation. A sprint must test the organisation's actual readiness instead of assuming an AI solution is appropriate.
Evidence. Gallup reported global manager engagement falling from 30% to 27% in 2024, while individual-contributor engagement remained at 18%.
Interpretation. The product should remove coordination work and clarify decisions; it should not sell a fictional promise that accountability disappears.
Evidence. Microsoft's 2025 telemetry analysis reports an average of 117 emails and 153 Teams messages per worker per weekday. The accompanying survey covered 31,000 knowledge workers in 31 markets.
Interpretation. One decision should have one durable brief, one owner, one next action and one review date rather than another message thread.
Evidence. In early April 2025, the UK Office for National Statistics reported economic uncertainty affecting turnover for 32% of trading businesses; among businesses with ten or more workers, labour cost was the most reported challenge at 42%.
Interpretation. Recommendations should preserve cash, quantify assumptions and favour reversible tests with stop rules.
One decision, one evidence pack and a stated deadline. Missing information is logged rather than silently guessed.
Options, evidence, assumptions, authority, risks, next test, stop rule and review date in a reusable written brief.
The default recommendation is reversible and avoids new fixed cost unless the evidence supports it.
Replace a CEO prepares and tests the decision. The authorised person approves any real-world commitment.
Independent research note. Replace a CEO and Billi O Ner Management are not affiliated with the organisations cited above. Sources are included for inspection; citation is not endorsement.