You're Underestimating Your Own ROI by 5x

We thought we had a solid ROI model. We’d done the math, built the framework, grounded everything in real usage data. The number was real and defensible.

Then we talked to a former nurse at Argentum. And she told us we were off by about five times.

Not because our math was wrong. Because we didn’t know what we didn’t know. We’d built the model from the outside looking in. We knew the costs visible in the data: management time saved by self-serve answers, agency fees avoided through improved retention, efficiency gains from faster information access. Real costs. Also the ones easy to see from the outside.

What we were missing were the costs only someone who’s actually worked a frontline healthcare shift would know to count.

The transactions that go wrong because information didn’t arrive in time. The operational events that never make it into an incident report but still cost something, a delay, a workaround, a near-miss that took two hours to sort out because the right person wasn’t reachable. The compliance exposures that don’t become audit findings until it’s too late, but were created the moment a procedure wasn’t documented correctly. Care decisions made on incomplete information that require follow-up, correction, or escalation.

None of that shows up in our data. It shows up in the experience of someone who spent years on the floor.

Her assessment: our estimate of what the product saves these communities was conservative by roughly five times. Not because we inflated anything, because we didn’t know what to include. We weren’t undervaluing the product. We were undervaluing the problem it solves.

We transcribed the conversation on the drive home, ran everything through Claude with all the context, and came back with a revised model grounded in the real operational cost picture of senior living. The output was significantly larger than what we’d been presenting.

The lesson isn’t to inflate numbers. It’s that ROI for communication infrastructure in frontline-heavy industries is systematically underestimated, because the people building the models are outside the problem. The people who lived it know what’s actually at stake.

Find your Deanna. Then build the real model.