Most AI proposals get stuck in CFO purgatory because the ROI math is vague. Here's the exact framework we use in every proposal — battle-tested across dozens of approvals.
Step 1: Quantify current cost. For each workflow you're automating, calculate (people affected) × (hours/week) × (loaded hourly cost) × 50 weeks. This is your annual current cost.
Step 2: Quantify post-automation cost. Estimate how many hours the automated workflow will still consume (usually 10–20% of original). Calculate the same way. The difference is your annual savings.
Step 3: Calculate payback period. Implementation cost ÷ annual savings × 12 = months to payback. Anything under 6 months gets approved fast. Under 3 months gets approved instantly.
Step 4: Add the second-order benefits — but conservatively. Faster cycle times often unlock revenue (more placements, more renewals retained, more clients served). Quantify these conservatively. CFOs respect conservative numbers more than aggressive ones.
Step 5: Stress-test it. Assume you only achieve 50% of the projected savings. Is the ROI still acceptable? If yes, you have a defensible business case.
Example. 5 recruiters × 15 hrs/week × $65/hr × 50 weeks = $243,750 current annual cost. Post-automation: 5 × 3 × $65 × 50 = $48,750. Annual savings: $195,000. Implementation: $18,000. Payback: 1.1 months. Approved.