July 14, 2026 · 6 min read · AI ROI · Business Automation

How to Calculate AI Automation ROI (With the Formula We Use)

Every automation vendor promises ROI; almost none show the math. Here's the actual formula we use in discovery to rank a client's automation opportunities — it's simple enough for a spreadsheet and honest enough to kill bad projects before they start.

The value side: four components

The cost side people forget

Build cost is visible. The forgotten lines: API and platform fees at your real volume (model tokens, per-task pricing), maintenance (integrations change, workflows need tuning — budget a monthly amount, not zero), and human review time where you deliberately keep checkpoints. An honest ROI model includes all three; a vendor model usually includes none.

The payback rule of thumb

We aim first at automations with payback under three months — typically high-frequency, rule-heavy processes like data entry, reporting and follow-up. Deeper systems like AI agents pay back bigger but slower; sequencing quick wins first funds the ambitious builds and earns internal trust. One caution from real projects: if a process is broken, automation makes it fail faster. Fix the process, then automate — that's why our consulting starts with process mapping, not tools.

Run your own numbers

Pick your three most repetitive processes, run the formula, and rank by payback. If the top one clears three months on conservative inputs, you have a business case — with numbers your CFO can interrogate, which beats any vendor's slide.

Keep reading

Ready to see what this looks like for your business?

Book a free strategy call — we'll map your highest-ROI automation opportunity, no obligation.

Prefer email? Contact us here.