The pitch is always the same: a little capital, a network, some mentorship, twelve weeks of structured pressure, and a demo day at the end. Most accelerators make that offer. A working paper published in April through the National Bureau of Economic Research, by Youn Baek and Deepak Hegde of NYU Stern, examined nearly 750,000 American startups and found that the average accelerator produces outcomes no better than going it alone, and often measurably worse. The mechanism isn't complicated. A bad programme pulls a founder out of their business for three months, surrounds them with generic advice, and sends them home with slightly less runway and a lot of new opinions.
What separates the programmes that work? According to the NYU Stern research, the differentiator is selectivity and specificity. Accelerators with rigorous intake processes, narrow sector focus, and direct access to paying customers or investors at the end, rather than just a room full of other startups, outperform the field by a significant margin. Cohort size matters too. Smaller groups where founders get genuine individual attention beat large-batch programmes where mentors are spread thin. The paper also found that equity taken by the accelerator in exchange for its services is a reliable signal: programmes that take little or no equity tend to be more serious about founder outcomes, because their reputation is the only return they get.
In Scotland, the landscape is better than the global average suggests. Scottish EDGE, now in its thirteenth year, operates a grant and loan model rather than an equity grab, and its alumni include some of the country's most recognised scale-ups. Techscaler, the Scottish Government's national tech accelerator launched in 2023 and delivered through the University of Strathclyde, has been designed specifically around the gaps the research identifies: sector-relevant mentorship, access to real customer networks, and no equity requirement. The Scottish Government's own economic development framework, backed by Scottish Enterprise and Highlands and Islands Enterprise, has pushed programme design in exactly the direction the NYU data recommends.
The Baek and Hegde research draws on a separate strand of evidence worth noting: founders who went through high-quality programmes reported that the most valuable thing was not the curriculum or the capital. It was warm introductions to people who could actually write a cheque or sign a purchase order. That finding maps closely onto what Business Gateway advisers in Edinburgh and Glasgow have long told early-stage founders: the right room matters more than the right workshop. A programme that can put a Scottish healthtech founder in front of NHS procurement, or a climate startup in front of a fund that actually backs climate, is doing something categorically different from one that runs a pitching workshop on a Tuesday morning.
The practical implication for any Scottish founder being courted by an accelerator right now is simple. Ask three questions before you commit. First: who specifically will you meet, and can you talk to a founder from a previous cohort who met them? Second: what equity, if any, does the programme take, and what is it actually spending on you? Third: has anyone from this programme's alumni raised a follow-on round or landed a significant contract in the last eighteen months? If the answers are vague, that is data. The research from NYU Stern is clear that a bad accelerator is not neutral, it is a cost: in time, in focus, and sometimes in morale. Scotland has good programmes. Make sure you're applying to one of them.