Stacking subsidies in practice: WBSO + MIT + SLIM on a single project

Key points: An SME wholesaler implemented an AI demand-forecasting system in 2025 for EUR 82,000. By smartly combining WBSO (EUR 11,700), MIT feasibility (EUR 7,000), MIT R&D collaboration (EUR 16,450) and SLIM (EUR 3,600), they secured EUR 38,750 (47%) in subsidies. Net cost: EUR 43,250. This article shows how this stack was built, with data, deadlines and pitfalls.

The project situation

Company: Wholesaler in technical components, 38 FTE, EUR 25M revenue, one DC in Apeldoorn. Problem: stock-outs during seasonal peaks + overstock during troughs = ~EUR 600k in lost margin per year. Solution: AI demand forecasting per SKU per week, fed by 4 years of sales history and external signals. Project size: EUR 82,000 (excl. own internal hours), 9-month lead time, in collaboration with a data partner.

The subsidy stack

1. WBSO (EUR 11,700)

When applied for: January 2025 (ahead of the Q2 WBSO period, April-June).

What it covers: labour costs of the two external developers (Stratalytic + data partner) invoiced in the Netherlands, plus the in-house data engineer's hours.

Calculation:

• R&D labour costs in scope: EUR 33,000 (data engineering, model development, integration)
• WBSO percentage: 36% (standard rate, non-starter)
• WBSO benefit: EUR 33,000 x 36% = EUR 11,880, rounded to EUR 11,700 after correction for non-awarded hours

Key detail: only development hours count. Project management, sales and marketing do not.

2. MIT feasibility (EUR 7,000)

When applied for: May 2025 (ahead of the June window).

What it covers: the first 3 months of the project - data audit, model testing, baseline comparison. This part is "research", not "production build".

Calculation:

• Project costs for the feasibility phase: EUR 20,000
• MIT percentage: 35% (feasibility)
• MIT feasibility benefit: EUR 20,000 x 35% = EUR 7,000 (well under the EUR 20,000 cap)

Key detail: MIT feasibility may overlap with WBSO but may not cover the same hours. The budget above splits this explicitly: WBSO for development hours, MIT for cloud compute, external expertise and evaluation.

3. MIT R&D collaboration (EUR 16,450)

When applied for: July 2025 (after a successful feasibility study).

What it covers: months 4-9 of the project - production build, collaboration with the data partner, joint development.

Calculation:

• Project costs for the R&D phase: EUR 47,000 (collaboration portion with the data partner)
• MIT R&D percentage: 35% of collaboration costs
• MIT R&D benefit: EUR 47,000 x 35% = EUR 16,450

Key detail: requires a collaboration agreement between SME partners. Stratalytic + the data partner signed an agreement setting out costs and ownership rights.

4. SLIM (EUR 3,600)

When applied for: September 2025 round.

What it covers: AI-literacy training for the procurement team + management who would work with the demand-forecasting dashboard. At the same time, EU AI Act compliance.

Calculation:

• Training costs: EUR 6,000 (in-company training, 8 employees, 2 days)
• SLIM percentage: 60%
• SLIM benefit: EUR 6,000 x 60% = EUR 3,600

Key detail: the training had to be delivered by a certified provider. Stratalytic combined this with our own trainers + an external AI Act specialist.

The total bill

(Note: the project amount of EUR 82,000 was the net cost without subsidy; gross including training came to EUR 106,000.)

The timeline (12 months)

What went wrong (and how it was solved)

1. MIT feasibility delay. Due to a backlog at RVO, the assessment took 7 weeks instead of 4. Workaround: WBSO was started in parallel so developers did not sit idle.

2. Data quality worse than expected. 18 of the 48 months of history had missing or inconsistent values. Workaround: the MIT feasibility scope was extended (within budget) to include data cleaning. This required a formal scope-change notification to RVO.

3. Missing the SLIM round. The first SLIM application in the March 2025 round was full; the budget was gone. Workaround: included in the September round - 6 months later. Not ideal, but applying for SLIM across two rounds in parallel does not work.

4. Collaboration ownership rights. The data partner initially wanted co-ownership of the model. Workaround: the collaboration agreement stipulated that the wholesaler becomes the owner; the data partner has a licence to use the algorithm in comparable projects.

Ten lessons from this project

1. Start with WBSO first. No competition, fast approval. An early win builds confidence.
2. MIT feasibility and MIT R&D are separate tracks. Apply sequentially, not at the same time.
3. Stack smartly, no double coverage. The budget must explicitly split costs between subsidies.
4. SLIM rounds are limited. Miss one, and you wait 6 months.
5. Collaboration agreements come before the MIT R&D application. No signature, no application.
6. Data quality is always worse than expected. Reserve 15-25% of the budget for data cleaning.
7. Schedule compliance up front. Carry out the EU AI Act assessment during feasibility, not at the end.
8. Do not underestimate your own founder time. ~150 hours over 12 months for stakeholder work.
9. Document everything from day one. RVO audits can come up to 5 years afterwards.
10. Plan for year 2 from day one. What is the follow-up subsidy if this works?

How Stratalytic would do this for your company

At Stratalytic, stacking is standard. For every client project we write the WBSO application, assess MIT options, and look at SLIM for the team. You get:

• One partner for both application and execution
• One project plan that covers all subsidies
• One invoice with clear subsidy allocation
• Support with final accountability towards RVO

Schedule a 30-minute introduction and we will look at which stack is possible for your project.

Related:

• Which AI subsidy fits me
• MIT vs WBSO vs SLIM comparison
• Stacking subsidies for maximum benefit

Frequently asked questions

How much can I stack at most? Dutch state-aid rules cap cumulation at 35-50% depending on company size. Realistic in practice: 35-45%.

Can I claim the same hours for both WBSO and MIT? No. WBSO covers labour costs, MIT covers other project costs. Double coverage is prohibited.

What if one subsidy is rejected? Then the rest still stands. Stacking is not all-or-nothing. Start with WBSO for certainty.