3.7x ROI per dollar invested, 5.1-month median payback, 44% of companies with positive ROI in 12 months (McKinsey 2026).
Direct gains: hours saved × hourly cost, error reduction × error cost, avoided tickets × ticket cost. Indirect gains: quality improvement (CSAT, NPS), time-to-market acceleration, risk reduction.
Licences/API + cloud infrastructure + development + training + maintenance + governance/compliance. Classic mistake: underestimating maintenance cost (typically 20-30% of initial cost per year).
Payback = Total investment ÷ Monthly net gains. NPV over 3 years at 10% discount rate. Require payback <18 months for medium-risk projects.
Define KPIs BEFORE deployment. Example: time-to-hire (baseline 45d → target 27d), false positive fraud rate (baseline 3% → target <0.5%), NPS (baseline 32 → target 40+).
Typical investment: €25,000-50,000
Annual saving: €8-15 × invoice volume
Payback: 4-8 months
3-year ROI: 4-8x
Typical investment: €30,000-80,000
Saving: (avoided or redeployed agents) × hourly cost
Payback: 6-12 months
3-year ROI: 3-5x
Investment: $19/month × developers
Gain: +25-55% code productivity
Payback: <1 month
3-year ROI: 10-15x
Investment: €40,000-100,000
Gain: reduced information search time + accelerated onboarding
Payback: 8-14 months
3-year ROI: 3-6x
McKinsey's 2026 benchmarks put the median ROI at 3.7x per dollar invested, with a median payback of 5.1 months. But only 44% of companies reach a positive ROI within 12 months: the median hides wide dispersion.
Compare the measurable gains (time saved, costs avoided, revenue generated) with the full cost (licences, inference, integration, human oversight). Set a baseline BEFORE deployment, otherwise the gain cannot be demonstrated.
Recurring causes: no baseline, a use case chosen for the hype rather than the value, underestimated inference costs, and a pilot that never reaches production.
Molderez Consult SRL co-builds your 18-month AI roadmap.
AI strategy workshop