How to Actually Calculate the ROI of AI in Your SME

Many SMEs want to introduce AI — but the business case is missing. Here is an honest, traceable framework for calculating ROI. With concrete figures from real projects.

Why Most AI ROI Calculations Fail

"AI increases productivity by 30%." SME owners hear this from consultants, vendors, and conference speakers. And it's true — somehow. But which 30%? Which employees? Which tasks?

The problem is abstraction. ROI calculations that work with percentage improvements in overall productivity are useless for decision-makers. What a managing director needs: which specific tasks consume how much time — and what will AI eliminate or accelerate?

The AIPE ROI Framework: 4 Cost Drivers

1. Information search: Employees spend an average of 14 hours per week searching for information. That is 35% of paid working time.

2. Onboarding costs: Reaching full productivity takes 3 to 6 months. An AI system that makes company knowledge immediately accessible can halve this. At €55,000 salary and 4 months reduced productivity, that is almost €18,000 per new hire.

3. Error costs from outdated information: Wrong prices, outdated specs, poor decisions due to missing context — regularly five-figure per incident in manufacturing.

4. Opportunity costs from slow decisions: When a request cannot be answered promptly, the company loses speed.

Worked Example: Machine Manufacturer With 80 Employees

20 employees in knowledge-intensive roles. Average hourly rate including overhead: €65.

Information search: 20 employees × 14 hours/week × €65/hour × 48 weeks = €870,000/year in search time. Conservative assumption: AIPE eliminates 40% = €348,000 savings.

Onboarding: 5 new hires/year × €18,000 productivity loss × 50% reduction = €45,000 savings.

Total savings (conservative): €393,000/year.

AIPE cost: Use the currently published plan and user count from our pricing page in your calculation.

ROI: Divide expected net savings by current total cost. This keeps the calculation auditable when plan or usage changes.

Realistic Expectations: What Customers Actually Achieve

Machine manufacturer (140 employees): −65% effort for shift documentation
Logistics provider (320 employees): −70% coordination effort for exceptions
Management consultancy (90 employees): 3× faster onboarding of new consultants
Typical payback period: 6–12 weeks after go-live

How to Build Your Own Business Case

Step 1: Time tracking for one week in three departments.

Step 2: Scale to annual basis.

Step 3: Take 30% of that sum as a conservative AI saving.

"The ROI was measurable after 6 weeks. I had expected it to take months." — Managing director, financial services, 60 employees