Someone pulls up the dashboard in a review. Nine of eleven teams on the internal platform, the line climbing quarter over quarter, the room nodding. Healthy adoption. Mark it a win.
The chart can’t answer the one question that matters: how many of those nine would still be there if they had somewhere else to go.
An external product earns every user. People had options, picked you, and churn the day something better shows up, so the adoption curve is a running tally of real choices. Each tick is information because the alternative was real.
An internal platform rarely works that way. A team is on it because leadership said standardize, or because building their own means defending duplicate spend in a review they’d rather skip, or because there’s no second option the org will fund twice. Their presence isn’t a judgment on the platform. It’s a judgment on their alternatives, and that tells you nothing about what you built.
So the same chart means opposite things depending on one fact: could the user have walked. High adoption in a free market is evidence. High adoption in a captive one is just the baseline. A mandated platform with miserable ergonomics and a genuinely good one post identical numbers, and the metric every platform team puts at the top of the deck can’t tell them apart.
The most captive consumer of all is the one you built it for. Your first app didn’t adopt the platform. It became the platform, and it can’t leave, because there’s nowhere to go and no separate team to do the leaving. The heaviest usage in your whole system is also the least informative number in it.
Captivity hides the defections too. A team that’s nominally on the platform but keeps a private script to redo the step you made painful still counts as adopted. Captive users don’t churn. They comply on paper and route around you in the details, and there’s no turnstile at the exit to count the ones who already left in spirit.
So stop reading usage as a verdict and go find the vote where it leaks out. Ask a captive team what they’d build instead, and listen for whether you get a shrug or a spec. Count the workarounds, not the logins. Watch the greenfield case: the one time a team starts something new with no mandate attached yet, do they reach for you or around you. A free choice on a new service is the closest thing to a market you get inside a company, and it’s where the truth gets out before the dashboard hears about it.
The teams that get blindsided aren’t the ones sitting at low adoption. Low adoption at least tells you something’s wrong while you can still fix it. It’s the ones at ninety-five percent, reading a captive room as a standing ovation, who never see the reorg coming that hands one of those teams an exit and watches it gone in a week.