What the Western Balkans Experiment Really Shows
The subject here is a specific randomized test of investment-readiness support, not a verdict on every incubator. Its value is that it separates an intensive offer from a cheaper alternative and then follows firms beyond the immediate programme. Reading it correctly helps answer the practical question: can a support intervention make a startup more prepared without necessarily making the startup grow faster? The evidence supports a clear short-run result, a more uncertain long-run result, and a sharp warning against treating either one as the other.
What the experiment was—and was not
The study examined 346 innovative small and medium-sized firms in Croatia, Kosovo, Macedonia, Montenegro, and Serbia. Eligible firms were randomly allocated to one of two groups: 174 were offered an intensive investment-readiness programme and 172 received access to a much cheaper online course. The groups were balanced on initial characteristics through the assignment process, so a later difference between them can be interpreted as the effect of being offered the intensive programme rather than as a simple difference between stronger and weaker applicants.
That last phrase is important. The experiment estimated an intention-to-treat effect: what happened because a firm was assigned to the intensive offer rather than the online alternative. It did not estimate the effect of a company’s decision to join an incubator compared with the decision of a similar company not to join one. Both groups received a form of support. The comparison was designed to answer a policy question that is more demanding than “does any help beat no help?”: is an expensive, intensive intervention necessary, or can a low-cost and more accessible online option produce similar results?
The intensive offer was a two-month programme built around work, not a badge. Firms used the WhatAVenture platform to describe the problem they addressed, their commercialisation concept, expected revenue streams, market size, and competitive position. A lead mentor provided feedback. The acceleration phase added individual mentoring from the lead mentor and a pool of more than 100 specialised mentors, four masterclass weekends, and a final period of pitch-deck development, practice pitches, and on-site preparation.
The control group received an online Innovation Readiness Series course. It contained ten modules of roughly 45 to 60 minutes, covering issues such as communicating the benefits of an innovation, intellectual-property protection, market validation, competition, and pitching. The estimated cost of the intensive treatment was about $4,000 per active participant, while the online course cost $153 per firm after its setup charge. Those figures are not a simple cost-effectiveness calculation: the treatment estimate varied with services actually used, and some shared advertising and event costs were excluded. They do, however, make the contrast concrete. The treatment bought substantially more individual attention and live interaction than the online alternative.
The study also shows why a bundle must not be mistaken for a single mechanism. The intensive group received platform work, mentoring, group instruction, and pitch preparation together. The experiment cannot tell us whether mentoring mattered more than masterclasses, whether the platform did most of the work, or whether the live pitch preparation was decisive. Nor did it test workspace, direct equity finance, internationalisation services, or the full ecosystem-hub model described by the OECD. It tested one investment-readiness package in one regional setting against one online comparison.
Take-up reinforces the same boundary. Of the firms assigned to the intensive programme, 90.1% completed the online platform and 79.3% received individual mentoring, but 43.7% attended at least one masterclass in person. Videos were available for those who could not attend.