A well-structured AI business case fits on one page. Here's the template we give clients who need to get sign-off from a skeptical CFO.
The hardest part of deploying AI isn't the technology. It's getting budget approved. And the reason most AI proposals get blocked isn't that the ROI isn't real — it's that nobody translated it into language a finance team trusts.
The one-page structure
-
01
The problem, in dollars.
"Our customer success team spends 280 hours/month on manual status updates. At a fully-loaded cost of $65/hr, that's $218k/year — of which at least 70% is non-judgment work."
-
02
The proposed solution, in one paragraph.
What the agent does. What it touches. Where the human stays in the loop. No jargon.
-
03
The investment, broken down.
One-time build cost. Annual run cost. Internal hours required. Sensitivity range.
-
04
The return, modeled conservatively.
Hours reclaimed × blended rate. Error reduction × cost per error. Cycle-time reduction × revenue impact. Conservative, not aspirational.
-
05
The payback period and confidence.
Months to break-even. Three scenarios (base / downside / upside).
-
06
The risk disclosure.
The things that could go wrong, and what you'd do about them. CFOs trust the person who brings them problems they hadn't thought of.
The fastest way to get blocked is to show up with "massive productivity gains." The fastest way to get approved is to show up with hours × dollars.
What convinces a CFO
Three things, in order: a clear unit of work (hours, tickets, matters), a conservative conversion rate, and a payback period under 12 months. If you have those three, most CFOs become champions. If you don't, most CFOs stay skeptical.
Book a 30-minute call
We'll work through the numbers with you. Walk out with a first-draft business case in your language.
Schedule it →