Look at startup support programmes across different countries and you’ll notice how similar they are. An accelerator modelled on Y Combinator or Techstars. A pitch competition. A seed grant scheme. A co-working space. A demo day.
These models keep being funded, year after year and country after country, even where results have been disappointing. That isn’t because the people designing them aren’t thoughtful. It’s because these models are explainable, measurable and repeatable, and those qualities matter a great deal inside government.
The governance challenge policymakers face
Policymakers working on innovation face real constraints. Public money has to be justified. Budgets run on annual cycles. Ministers need results they can report, often within a single term of office. Every programme has to survive audits, parliamentary questions and changes of leadership.
In that environment, a programme that’s easy to explain and easy to measure has a big advantage over one that might work better but is harder to describe. “We supported 50 startups and held 12 events” is a clear sentence for an annual report. “We improved the decision-making capability of 40 founders” is much harder to put on a slide.
Why visible metrics dominate
The metrics most startup programmes report are visible and countable: startups supported, participants trained, events delivered, funds disbursed, jobs announced. They can be measured within a budget cycle, and they go up when more money goes in.
Founder capability, by contrast, is mostly invisible to these systems. The change in how a founder interprets evidence, makes decisions under pressure or responds to failure doesn’t show up in a standard report, even though it’s what ultimately decides outcomes.
So programmes optimise for what’s measured. It’s a familiar pattern, often described as Goodhart’s law: when a measure becomes a target, it stops being a good measure. Programmes become very good at producing startups supported and events held, without necessarily producing better companies.
How intervention models become entrenched
Once a programme model is established, several forces keep it in place.
Templates come with ready-made metrics. An accelerator model arrives with a well-understood structure and a standard set of KPIs. That makes it easy to procure, easy to manage and easy to evaluate.
Familiar models feel lower-risk. Procurement and approval processes naturally favour approaches that have been funded before, elsewhere. Choosing a familiar model is easier to defend than choosing a different one, even if the familiar one hasn’t worked.
Success is defined by the model itself. If a programme is judged on the number of cohorts delivered, then delivering cohorts counts as success, whatever happens to the companies afterwards.
The result is a cycle. Programmes are judged on activity, activity goes up, the programme is declared a success and funded again. Activity is up. Outcomes aren’t.
The cost of the cycle
This cycle has a real cost. Despite billions in government investment, value creation across startup ecosystems remains extremely uneven. Startup Genome’s 2023 analysis puts North American ecosystem value at roughly $5.3 trillion, against $281 billion for MENA, $197 billion for Latin America and $24 billion for Sub-Saharan Africa over the same period.
Many of the models being funded were designed for very different conditions. Accelerators in Silicon Valley work with founders who are often on their second or third company, in an ecosystem full of “elders” who have built and scaled companies before. In most emerging ecosystems, close to 100% of founders are first-timers, and there are few local elders to guide them.
Importing the model without those conditions means taking founders who are already working on an idea and helping them move faster, whether or not their direction is right. The accelerator teaches them to do the wrong thing better, rather than teaching them to do the right thing well. The activity metrics still look good. The outcomes don’t change.
What an honest definition of success would look like
An honest definition of success starts with a simple question: are we producing founders who can build companies that grow?
That points to a different set of measures:
- Founder capability: how founders’ mindset, knowledge, skills and patterns of behaviour change over the programme, as observed by experienced mentors
- Decision quality: whether founders validate problems, consider alternatives and act on evidence, assessed before results arrive
- Company progress: survival, customer traction and follow-on investment over several years, not just at demo day
- Repeat founders: how many alumni go on to build second and third companies
- Local elders: how many alumni become mentors, investors and early employees for the next generation
Some of these take longer to show. But capability and decision quality can be assessed within a single programme, giving policymakers credible evidence well inside a budget cycle.
It also means being honest about what’s possible. At Global Innovation Dojo, we don’t promise unicorns. We don’t sell theatre. We don’t inflate claims.
A practical path for policymakers
Changing approach doesn’t have to mean taking a big risk. There’s a sensible progression that fits how governments work:
- Start with an entrepreneurship discovery. A short diagnostic of the ecosystem against the conditions founders actually need (mentorship density, funding access, regulatory support and cultural attitudes to risk) gives an evidence-based starting point before any programme is committed.
- Test with a founder mindset bootcamp. A one- to four-week immersion is the fastest, lowest-cost way to see whether founder development belongs in the ecosystem strategy.
- Commit to a year-long cohort. An annual programme that takes a defined group of founders from validating an opportunity to a repeatable go-to-market delivers a complete, measurable outcome within a single budget cycle.
- Build a long-term partnership. Over time, develop the full founder pipeline, and grow the local mentors and elders the ecosystem lacks, treating founder development as national infrastructure rather than a one-off programme.
Impact depends on whether effort is applied at the real constraint
Governments don’t keep funding the same models because they’re unaware of the problem. They fund them because the system rewards what can be explained and counted.
The opportunity is to count different things, and to point investment at the real constraint: the capability of the founders making the decisions. That’s the highest-leverage investment a government can make in its startup ecosystem.
Want to explore a founder-first approach for your ecosystem? Talk to us.