Here is a contradiction worth sitting with. Sturdy will tell you that a customer with high usage might be your biggest churn risk — opening support tickets, emailing their CSM, hammering the product not because they love it but because they can't figure something out. Matt Tidwell will tell you the opposite failure is, per Tidwell, just as fatal: a customer quietly achieving a real outcome that nobody ever captures. Both are looking at the same dashboard. Both dashboards are green. Both customers are gone by renewal.
That is the whole problem in one frame. Every signal your product can detect on its own — logins, DAU, sessions, feature adoption — measures activity, and activity is a leading indicator of nothing. The variable that actually decides renewals doesn't live in your telemetry. It lives in your customer's head.
This week the diagnosis has produced named, publishable cures — reportedly three or four frameworks landed in the same window, all circling the same blind spot, all making the same radical move.
Start with the scene that named the paradox. Gaurav Sharma watched a customer whose usage dashboards were green across the board nearly churn at renewal. Every metric his team trusted was measuring the wrong thing. Out of that near-miss he built the Outcome Clarity Test — a way to close the gap between what ships and what actually changes for the customer.
The reason this keeps happening isn't incompetence. It's asymmetry. Engagement is measured continuously — it streams into your dashboard every second a user is logged in. Outcomes are measured never, because your product cannot detect them. It can see that someone ran a report. It cannot see whether that report saved the customer real money or got quietly ignored.
Tidwell's Retention Architecture makes this the load-bearing insight of the whole system. He maps six zones where retention breaks, and the one enterprise teams miss sits late in the sequence. Realization is where the customer has to prove the product is working — "usage metrics alone can't do this," he writes. Then comes Recognition, the zone he reportedly describes as often invisible: the customer achieves a meaningful outcome, but it's never captured, celebrated, or communicated.
His line is the one to tape to your monitor: "Unrecognized value is functionally the same as unrealized value."
Unrecognized value is functionally the same as unrealized value. A customer can be genuinely winning with your product and still churn.
Read that twice. It means a customer can be genuinely winning with your product and still churn — because the win never got written down, the champion who felt it moved to a new role, and at renewal there is no story to tell procurement. "We use it a lot" is not a defensible answer. The value existed. It just couldn't be spoken.
This is where the frameworks converge on their one shared move. They all stop trying to detect value and start trying to make the customer articulate it. They redefine churn and value as things a customer must be able to say out loud, not things the product can infer.
Sturdy makes the case that raw usage is "riddled with false positives," and its fix is to track contextual signals telemetry can't produce: budget and procurement discussions, expansion or contraction language, change-order requests, closed-loop responsiveness. Notice what every one of those is. It's a sentence a human said. It's language, not clicks.
Sturdy also rescues the inverse case, the one that gets healthy customers accidentally flagged as risks. Low usage, they argue, can mean the customer finished implementation and is now deriving value without needing to log in. "If your product makes life easier, customers shouldn't need to use it constantly." Punish that customer with a re-engagement campaign and you're solving a problem that doesn't exist while ignoring the frustrated power user next door.
Kelvin Obasuyi hit the same wall from inside a scaling company. As Duplo pushed past 2,000 merchants, he found four retention metrics that standard dashboards were actively hiding. The lesson generalizes: the metrics that predict retention are usually the ones your tooling wasn't built to capture, because they were never events in the first place.
You can even watch this disease one layer down, inside the models. A relevance PM describes systems where ranking precision improved month over month while users kept saying the experience was "not quite right" — offline metrics healthy, actual outcome poor, because the system learned to represent content faster than it learned user intent. Same pathology, different altitude: if your evaluator can't verify the outcome, no downstream dashboard can either.
Now the turn, and it's the part most teams underestimate.
Once you accept that outcomes live in the customer's head, retention stops being purely an analytics problem. Building the outcome model is the easy half. The hard half is that outcome measurement requires the customer to articulate value — which makes this a language and evidence problem.
Tidwell is explicit about what this demands: internal champions who can tell the story of how value is being delivered, and "those champions need data, language, and frameworks to make that case inside their organization." Your job is no longer just ensuring the customer uses the product. It's ensuring the customer can prove the product works — to their own boss, in a room you're not in.
That reframes the entire CS function. Reactive ticket-handling and quarterly check-ins don't produce articulated value. They produce activity logs. What produces articulated value is deliberately manufacturing the evidence and handing your champion the words to use it.
So here's the operating implication, and it's blunt.
Stop asking "is usage healthy?" Replace it with a harder question: can the champion prove ROI without us in the room? That is the variable that actually predicts the renewal, because it's the exact test procurement will run when the contract comes up.
Then instrument the answer weekly. Not at renewal — that's the autopsy. Every week you should be able to name, for each meaningful account, the specific outcome the customer achieved, whether anyone on their side captured it, and whether your champion could recite the ROI story cold. If any of those three is blank, the account is at risk regardless of what DAU says.
This is why the wave arrived all at once. Sharma's Outcome Clarity Test, Tidwell's Realization and Recognition zones, Obasuyi's four hidden Duplo metrics, Sturdy's context signals — they're four attempts to instrument the same variable that has always sat outside the product's field of view.
The uncomfortable truth underneath all of them: your dashboard was never lying about engagement. It was telling the truth about the one thing that doesn't determine whether you keep the customer.
The customer keeps you when they can say, in their own words, what you did for them. Your only real job is making sure they can — before the renewal call, not during it.