ROI Calculation for AI Projects

Without a clear ROI calculation, AI projects fail at budget meetings. This practical template helps you create the business case for any AI initiative.

The ROI Framework in 5 Steps

Step 1: Capture Costs

Cost Block	One-time	Recurring (p.a.)
Software/licenses	€20,000–100,000	€12,000–60,000
Integration & setup	€30,000–150,000	—
Training	€5,000–20,000	€2,000–5,000
Data preparation	€10,000–50,000	€5,000–15,000
Personnel (internal)	—	€40,000–80,000

Step 2: Quantify Savings

Direct savings:

Time saved × hourly rate × number of employees
Reduced error costs (complaints, rework)
Saved external services

Indirect savings:

Faster time-to-market
Higher customer satisfaction → lower churn rate
Better decision quality

Step 3: Calculate ROI

ROI = (Annual Benefit - Annual Cost) / Total Investment × 100

Example:
Annual benefit:       €180,000
Annual cost:           €60,000
Total investment:     €120,000

ROI = (180,000 - 60,000) / 120,000 × 100 = 100%

Step 4: Payback Period

Payback = Total Investment / (Monthly Benefit - Monthly Cost)

Example: 120,000 / (15,000 - 5,000) = 12 months

Step 5: Risk Assessment

Multiply expected benefit by a probability of success:

Conservative (60%): For initial business cases
Realistic (80%): After successful pilot
Optimistic (100%): Only for proven use cases

Practical Tips

Always calculate conservatively — positive surprises beat disappointments
Show indirect benefits separately — they convince but are harder to measure
Document the comparison baseline — without a baseline, no ROI proof

Template download: Use this framework as an Excel template for your next AI business case.