What Slovenia's women's entrepreneurship programme taught us about paying the grant last

Between 2016 and 2019 Slovenia ran an entrepreneurship training programme for unemployed women with a degree, and released the EUR 5 000 start-up grant only once they had completed 100 hours of it. About 1 500 women took part. Two years after finishing, 90% were running an active business. The design choice that carried the result was the order in which the money moved.

Gate the money behind the training, not the other way round

Most start-up support pays first and trains second, because paying is easy to administer and training is not. Slovenia inverted the sequence. The certificate a participant earned at the end of 100 hours was the document that unlocked the cash, and the mentoring continued after the money landed. That did three things at once. It filtered for commitment without an assessment panel. It put the grant in the hands of a founder who already knew what to spend it on. And it gave the programme a natural checkpoint to measure. For any team running a grant-plus-training scheme where the two halves are only loosely coupled, this is the cheapest change available.

The problem was measured before it was addressed

Slovenia's government did not start from a hunch. A Global Entrepreneurship Monitor survey ranked the country among those with the lowest participation of women in early-stage entrepreneurial activity, and the government introduced a measure in response. The Promotion of Women Entrepreneurship programme, Podjetnost je zenskega spola, sat inside that measure. The targeting was narrow on purpose: women with tertiary education who had been unemployed for at least three months. That is a specific group, educated, formally out of work, and past the point where a job was likely to arrive on its own. Narrow eligibility is usually treated as a political cost. Here it was the thing that let a modest budget produce a coherent cohort.

What the programme did, and why it worked

The programme was a joint initiative between the public entrepreneurship and business development agency SPIRIT Slovenia and three other public bodies: the Ministry of Labour, Family, Social Affairs and Equal Opportunities, the Ministry of Economic Development and Technology, and the Public Employment Service of Slovenia. It ran in five cities, Ljubljana, Maribor, Celje, Nova Gorica and Novo Mesto, with two intakes a year in spring and autumn. Delivery was contracted out rather than built in-house: the Chamber of Commerce and Industry of Slovenia ran the training with local partners and mentors.

The certificate was the payment trigger

Participants took an introductory workshop followed by 15 in-person training sessions covering idea development, business modelling and legal advice. The sessions also walked participants through the public incentives already available to them, which quietly solves a real problem, because educated unemployed people frequently do not know what they are entitled to. On completing the 100 hours a participant received a certificate, presented it to the Public Employment Service of Slovenia, and was paid a one-time lump-sum start-up subsidy of EUR 5 000. Read that as an operating design rather than a bureaucratic quirk. The training provider assessed completion, a separate agency held the money, and the certificate was the interface between them. Neither body had to trust the other's judgement about the founder. They only had to trust a document, which is far easier to audit and far harder to argue about.

The support did not stop when the money moved

Follow-up training and mentorship remained available after the disbursement. Most schemes treat payment as the end of the relationship, which is precisely backwards. The first spend of a EUR 5 000 grant is where most of the avoidable mistakes happen, and it is the moment a new founder has the most questions and the fewest people to ask.

EUR 1 million in funding was made available for the training programme in 2018 and 2019. Across the full 2016 to 2019 run, about 1 500 women participated, and 90% had active businesses two years after completing. Set the grant against the training budget and the ratio is informative: the cash transfer is a large share of the cost per head, but it is the component the programme spends last, and only on people who have already invested 100 hours.

Transferable lessons

Five lessons travel well from Slovenia to other enterprise and employment support programmes.

• Gate the money behind the training. A completion certificate is a cheap, auditable trigger. It filters for commitment without an assessment panel and puts the cash with someone who has already built the plan it funds.
• Use a document as the handoff between agencies. When the trainer and the payer are different organisations, define one artefact that passes between them. It removes the need for either body to second-guess the other and makes the audit trail obvious.
• Narrow the eligibility rather than the ambition. Tertiary-educated women unemployed for three months or more is a specific population with a specific blockage. A specific cohort makes the curriculum easier to write and the results easier to read.
• Keep mentoring running past the disbursement. The riskiest weeks in a grant-funded business are the ones immediately after the money arrives, which is exactly when most programmes disengage.
• Fix a delivery cadence and hold it. Two intakes a year across five cities gave partners something to plan against and produced four comparable cohorts of data before anyone had to defend the budget.

What to watch out for

The design works where the training has real content and the target group can afford to attend it unpaid. Slovenia's participants were graduates receiving unemployment support, so 100 hours was possible for them. Apply the same gate to a group that has to choose between the training room and paid hours and the filter stops measuring commitment and starts measuring spare time. The headline result also needs handling with care. Selecting educated women who chose to complete 100 hours produces a cohort that would have outperformed the average regardless. The published figures record what happened to participants; they do not compare them against similar women who did not take part. Treat 90% as evidence that the sequencing did not harm outcomes and that the cohort was well chosen, not as proof of what the training alone caused.

Conclusion

Slovenia spent EUR 1 million a year on training and a EUR 5 000 grant per completing founder, and two years later 90% of about 1 500 participants were still trading. The instrument was ordinary. The sequence was not. Designing that sequence, and building the record that proves it worked, is the part most teams find hardest to resource on their own.