Do Startup Programmes Actually Commercialise Science?

Startup programmes such as incubators, accelerators, venture builders, and spinout support initiatives are widely presented as mechanisms for translating science into market outcomes. Yet in deep tech and science-based ventures, it remains unclear whether these structures are actually associated with hard commercialisation outcomes — or whether they primarily improve earlier-stage proxy signals such as startup formation, fundraising, patents, pilot activity, or short-term revenue. This paper asks a narrower and more consequential question: among European deep tech and science-based companies that have reached meaningful commercialisation thresholds, how visible is programme participation in their development path? The paper adopts a stricter commercialisation definition than is typical in programme reporting, using at least one of the following criteria: technology at TRL 8/9 or comparable industrial deployment evidence, recurring revenues above a defined threshold, IPO, or M&A. It combines a review of the existing literature on programme effectiveness with a preliminary coded sample of European companies meeting at least one of these outcome conditions. The methodological challenge is central to the analysis: programme participation is not systematically documented, and most existing studies do not measure long-horizon commercialisation outcomes directly. The preliminary evidence suggests a significant measurement gap. Existing programme research captures early-stage venture performance more reliably than hard science commercialisation, while the pilot sample indicates that many successful firms are more visibly associated with spinout origins, strategic partners, licensing models, or revenue-first commercialisation paths than with classical accelerator participation. The paper argues that current evaluation systems may overstate programme contribution to commercialisation because they rarely measure commercialisation itself.