Engaging isn't the same as valuable.

Engagement is the easiest thing in your product to measure and one of the easiest to misread. Time in app, sessions per week, taps per session, days active — every one of these counts attention captured. None of them counts whether the person got the thing they opened the app to get. Most of the time the two rhyme, which is exactly what makes the gap dangerous: you can go a long way trusting engagement as a stand-in for value before you notice they’ve come apart.

They come apart in both directions, and both are instructive.

A product can be intensely engaging and deliver almost nothing. The infinite feed, the streak you keep alive out of dread, the notification that pulls you back to a place you didn’t actually need to go — engagement soaring, value flat or negative. You built a better casino, not a better tool. And a product can be barely engaging and enormously valuable: the thing you open once a month that quietly saves you a day, the export button, the tool so efficient the best outcome is you close it and get on with your life. By an engagement dashboard, that second product looks like it’s dying. By any honest measure of value, it’s the one worth building.

The reason this matters for growth specifically: engagement is a leading indicator only when it’s engagement with the valuable action. Time spent finishing the job is a great signal. Time spent because you made it hard to leave is a signal pointing the opposite way, and it reads identically on the dashboard. When a growth team optimizes “engagement” in the abstract — more sessions, more time, more taps — it will, given enough quarters, find the changes that raise the number without doing the job. Not out of cynicism. Because those changes are easier to find than real value, and the metric can’t tell the difference. The number will reward you for the wrong thing with a completely straight face.

So the discipline is to name a value-delivered metric and keep it next to the engagement one. Not sessions — jobs completed. Not time in app — time saved, or the outcome the user came for, actually reached. Then watch what happens when the two diverge, because they will. A change that lifts engagement and lifts value: ship it, that’s the flywheel. A change that lifts engagement while value stays flat or drops: that’s the tell. You’ve probably found a hook, not a feature, and the honest move is to trust the value number over the engagement number even though the engagement number is the one that looks good in the update.

The blunt test to keep in your pocket: if a change made the number go up and you can’t say what job it helped someone finish, be suspicious of the number. Engagement is what it looks like when people can’t put your product down. Value is what it looks like when they’re glad they picked it up. You want both — but when you can only have one, the second is the business.

Worth reading around this: vanity metrics vs. actionable metrics; “jobs to be done” (the outcome the user hired the product for); the distinction between engagement as a goal and engagement as a proxy.