Dear Ryan,
I’m the Chief Technology Officer at a mid-size RIA. Eighteen months ago our board gave us the green light to begin investing seriously in AI. This opened the door for new tools, modernized infrastructure, dedicated headcount, outside consultants and more. It’s the kind of investment that shows up as a line item that makes investors lean forward in their chairs.
We’ve made real progress. Genuinely. Our advisors are spending less time on meeting prep. Our compliance review process is faster. We’ve automated workflows that used to eat up hours every week. The people closest to the work know something has changed. I know something has changed.
But here’s where I’m stuck: Our board wants a formal ROI analysis on our AI investment before they’ll approve next year’s budget. A number. Preferably a ratio. Invested X, returned Y, here is the multiple.
I’ve spent the last six weeks trying to build that model, and I keep running into the same wall. I think hours saved is a superficial metric. How do I calculate the ROI of an advisor who is less mentally exhausted at 4 p.m.? How do I put a number on a compliance process that didn’t fail? How do I quantify the institutional knowledge we’ve stopped losing every time someone leaves?
Claude and I are good with Excel, I can build a spreadsheet that tells a story. But I’d be telling a story, not measuring a reality or demonstrating real impact. And something about walking into that boardroom with a number I reverse-engineered feels gross.
Is the board asking the wrong question? And if they are — how do I tell them that without it sounding like I’m hiding a failed investment behind a lecture? That wouldn’t go well for me.
— Trying To Measure the Immeasurable
Dear Trying To Measure The Immeasurable,
Your board isn’t asking the wrong question because they’re unsophisticated. They’re asking it because nobody has given them a better one yet. That’s your job and it’s actually a bigger opportunity than you’re treating it as right now. Don’t miss this.
That’s your job and it’s actually a bigger opportunity than you’re treating it as right now.
The challenge with AI ROI as a unit of measurement is foundational: Technology doesn’t produce returns. Strategy does. The AI tools you’ve deployed are expressions of a strategic direction (faster advisors, tighter compliance, retained institutional knowledge, upskilling workforce). Measuring the tool’s ROI in isolation is like measuring the ROI of a highway without asking where it took you. The road didn’t create value. What moved across it did.
So put your Claude pencil down and listen to what you should build.
Go back to the strategic outcomes your AI investment was supposed to serve. Not the tools — the outcomes. Increase advisor capacity. Maintain compliance integrity. Retain premium talent. Elevate the client experience. Those are measurable. Maybe not perfectly, maybe not with a clean multiple, but directionally and honestly. Then show the board how the AI investment is moving those needles. Most importantly, don’t stop there. Illustrate what it would cost to move them any other way.
Less lecture, more strategy review. Your board is much more comfortable there … as you will be, too.
I’d walk into that room and confidently state: “You asked for the ROI on our AI investment. What I’m about to show you is more useful: the ROI on the strategy it’s serving. That’s the number that tells you whether we keep going and how fast.”
That reframe does two things simultaneously. It answers the accountability question they care about. And it quietly corrects the unit of measurement without making anyone feel wrong for asking.
Your board doesn’t need a ratio. They need confidence that the investment has direction.
Your board doesn’t need a ratio. They need confidence that the investment has direction. Those are completely different asks, and only one of them is answerable honestly.
Give them the actionable one.
— Ryan
Ryan George is the Chief of Staff of Docupace.
