Key Takeaways
A rigorous ROI framework for AI automation accounts for direct cost savings, revenue impact, implementation costs, failure rate adjustment, and opportunity cost. Formula: ROI = ((direct savings + revenue impact) × (1 - failure rate)) - (implementation cost + ongoing cost + opportunity cost)) / (implementation cost + opportunity cost). Applied across 47 client engagements, average first-year ROI was 310%. Median ROI was 180% — meaningfully lower, indicating outliers skew averages upward. This framework is designed for CFOs and operations leaders making build-vs-buy decisions about AI automation.
Why Most AI ROI Calculators Lie
Vendor ROI calculators are marketing tools. They include the savings and exclude the costs. They ignore failed implementations. They don't account for the 3–6 months where productivity dips while systems are being built. A real ROI calculation is harder and less impressive — but it's the one that will survive scrutiny from your CFO and board.
The Components of a Legitimate ROI Model
Benefits: (1) Direct labour savings — hours recovered × fully-loaded hourly cost. (2) Error reduction value — cost of errors prevented (rework, customer complaints, regulatory risk). (3) Revenue impact — pipeline increased or conversion rates improved attributable to automation. Costs: (4) Implementation cost — agency fees, internal time, infrastructure setup. (5) Ongoing operational cost — AI licensing, maintenance, oversight hours. (6) Failure rate adjustment — across the industry, 30–40% of AI implementations don't deliver projected results. Apply a 35% haircut to projected benefits.
The Formula
ROI = ((Direct Savings + Revenue Impact) × 0.65) - (Implementation Cost + Ongoing Cost × 12)) / (Implementation Cost). Example: £200k projected savings × 0.65 = £130k adjusted benefit. Implementation: £40k. Year 1 ongoing: £14.4k (£1.2k/month). ROI = (£130k - £54.4k) / £40k = 189%. This is honest. Agencies promising 400%+ ROI in year one are using vendor-grade assumptions.
Benchmarks From 47 Engagements
Average first-year ROI: 310%. Median: 180%. The gap matters — a small number of high-performing engagements inflate the average significantly. Engagements that hit >300% ROI in year one shared three characteristics: clean pre-existing data infrastructure, a dedicated internal owner (not a committee), and clearly defined success metrics before kickoff. Engagements below 100% ROI shared one characteristic: at least one of these three was missing.
Ready to implement this in your business?
Book a free AI Audit. 90 minutes. We'll map your highest-value opportunities and hand you a prioritised implementation plan.
Book My AI Audit