Are ecosystem leaders wasting money on accelerators?

• 

Startup ecosystems around the world are spending significant public money on accelerators and growth programs — but much of it isn't delivering. Not because of a shortage of initiatives, but because they are the wrong programs, run by the wrong operators, measured by the wrong metrics, designed for ecosystems that have fundamentally changed

The uncomfortable truth is that many program portfolios were designed for ecosystems that no longer exist. As AI-Native startups reshape every sector and Deep Tech moves from fringe to mainstream, generic accelerators continue to churn through cohorts optimized for a previous era. Even well-designed portfolios decay: ecosystems mature, founder needs evolve, new sectors emerge. Portfolios that aren't actively managed don't just stagnate, but actively misallocate resources, crowding out the programs an ecosystem actually needs. 

The wrong programs, measured the wrong way

A high-performing ecosystem needs a coherent support continuum from venture sprints and hackathons that broaden entrepreneurial participation, through pre-acceleration and core acceleration, to late-stage growth, commercialization, and global market access. Each stage requires different design logic, different operators, and different success metrics. Few ecosystems have a coherent framework across this full spectrum – and where the private sector does not sufficiently cover all critical stages, public support often has redundancy at some, critical gaps at others, and public budgets spread too thin to make meaningful impact anywhere. 

Governance failures compound all of this. Public agencies often operate programs directly, despite lacking the incentives, networks, and specialist expertise that high-performance acceleration demands. The consequences are predictable: inability to attract top-tier founders, weak mentor networks, susceptibility to political inconsistencies, and no mechanism to reward performance or remove underperforming operators. Competitive tenders for private sector operators — with genuine skin in the game via equity stakes or performance-linked funding consistently outperform direct public delivery, even if those private sector operators need to be imported from overseas damaging of all is how success is measured.

Perhaps most damaging of all is how success is measured. Many programs are still routinely evaluated on throughput — how many startups completed a cohort — rather than the metrics that actually matter: funding outcomes, revenue growth and global reach. And where these metrics are measured, there is often little attention paid to programs’ value-add: if programs are selective, it is simply not adequate to compare the selected startups with other, general startups without taking that selection effect into account. Without robust evaluation methodology, including longitudinal tracking against proper control groups, public funders have no reliable way to distinguish programs that add genuine value from those that merely select the most promising startups and take the credit. 

However, the research is clear that well-designed accelerator programs can indeed add value, both to startups and the founders themselves. Moreover, there can be significant ecosystem-level benefits: previous research has shown that venture capital investment increases across an ecosystem – even outside accelerated startups – when the first accelerators appear (see chart). These spillovers provide a clear rationale for public policy support. 

The Arrival of an Accelerator in an Ecosystem Increases Tech Investment

Source: Bone, Jonathan, Juanita Gonzalez-Uribe, Christopher Haley, and Henry Lahr. 2019. The Impact of Business Accelerators and Incubators in the UK (BEIS Research Paper Number 2019/009). BEIS. The chart shows average VC funding at local authority level in the U.K., before and after the first accelerator in that area.

What ecosystem leaders should actually do

First, they should start with an honest assessment of where their ecosystem actually stands – its strengths and weaknesses – and leaders’ objectives. A small, slow-growing ecosystem has fundamentally different priorities than a maturing one; the former must focus on increased startup creation before anything else. Startup Genome’s Lifecycle Assessment can assist this process.

From that baseline assessment, ecosystem leaders should consider their budget and what support is needed at each startup stage, from entrepreneurship activation through to global growth. In rapidly-developing ecosystems, it is also important to think ahead to the startup mix that will exist in 2-3 years. 

Policymakers should consider the strategic sub-sectors that matter most, and consult local industry about alignment. Specialization is no longer just a ‘nice-to-have:’ specialist programs consistently attract stronger applicant pools and better mentor-investor fit and, for some ecosystems, such as those with a clear industrial base in the sector concerned, can also play a role in startup attraction. Our Innovation Edge tool can help here. 

Next, policymakers should enable private sector operators, not compete with them, using expressions of interest to surface operators, and advertising beyond the local ecosystem where local suppliers do not exist. Financial sustainability is difficult, especially for programs providing very early stage support, and grants, tax reliefs, or public-private partnerships can all help, while public sector competition will harm. 

Many accelerators are like startups themselves: searching for a repeatable and sustainable business model. Ironically, however, the owners and operators sometimes fail to reflect the advice given to their startups and find themselves too consumed by day-to-day operations to think about their own strategy and productivity. Non-financial assistance – such as exposure to international good practice and smarter operating systems – may help in this regard, and there are now a number of tools which reduce the administrative burden for program managers and support them in making better data-led decisions. 

The eventual aim should be to develop a series of standalone private-sector program operators. However, to achieve sustainability, programs themselves may need to double-down on quality: increasing stipends or investment will attract higher-quality startups and better mentors, and increase the likelihood of equity-based programs making sizable returns. Public sector funding can also support this initially. 

Additionally, evaluation and accountability should be designed-in from day one, not as an afterthought. For example, it is much easier for programs to produce robust impact analyses if they track all applicants, not just the ones successfully admitted. Outcomes should be rigorously tracked, and public funders should be willing to expand what works, prune what doesn't, and add specialisms as their ecosystem evolves. Active portfolio management is not a one-time exercise, but an ongoing discipline. 

Accelerators and related growth programs can add substantial value – to founders, to their firms and to ecosystems as a whole. However, many public funders are not maximizing this value, through ill-defined strategy, portfolios that are inappropriate for their ecosystem, or weak program assessment.