A Decade of Effort, and Retention Hasn't Moved

Here is a fact that should bother the entire industry more than it does. Across a decade, SaaS retention — measured across the field — has essentially not improved. Not the tooling, not the QBR cadences, not the health-score platforms, not the enormous headcount the industry poured into Customer Success. The line is flat. Ten years of collective effort, and the number sits about where it started.

I’ve spent this whole series arguing why. When an entire field runs this hard at a problem and the needle doesn’t move, the honest conclusion isn’t “try harder.” It’s that the field has been pushing on the wrong thing.

SaaS retention has been flat for a decade. Every playbook, QBR cadence, health-score platform, and CS hire the industry bought in ten years — and the curve hasn’t moved. If the whole field is running hard and the number won’t budge, the field is pushing on the wrong thing.

To see what it should be pushing on, look at the actual shape of retention over time — not the one-year snapshot, the whole curve. A cohort of customers doesn’t decay to zero, and it doesn’t hold flat either. It falls — fast at first — and then, at some point, it levels off. A share of that cohort simply stops leaving. The curve stops dropping and runs flat for years.

That plateau is the most important feature on the chart, because its height is the whole story. The customers above the plateau line were always going to leave — they were renting, and they left on schedule. The customers at the plateau are the ones who bonded. The height of the plateau is your bonded fraction. It’s not a metaphor; it’s a number you can read straight off the shape.

And across the study, that plateau tends to settle at roughly a third of the original cohort — with enormous spread around it, from near-zero to the majority, depending on the two things I’ve spent a month on.

Retention doesn’t decay forever and it doesn’t hold forever — it falls, then levels off, and where it levels off is roughly a third of the original cohort. That plateau height IS your bonded fraction. Everything above the line was renting; everything at it is bonded.

So put the whole series in one frame. The plateau’s height is set by two things and only two things: whether you sold to the customers the product actually fits, and whether those customers reached a real result. Fit and results. Move either, and the floor rises. Ignore them, and no amount of dashboard-watching lifts it a single point.

Which explains the decade of flat. The industry spent ten years optimizing the three metrics — satisfaction, health scores, usage — none of which touch the height of the plateau. Meanwhile the two levers that do set it went essentially unmanaged, because almost nobody was measuring the plateau at all. You cannot improve a number you’ve never looked at, and the entire field has been steering by the wrong gauges. A decade of frozen retention isn’t a motivation problem. It’s a measurement problem.

A decade of frozen retention isn’t a motivation problem, it’s a measurement problem. Teams optimized satisfaction, health, and usage — none of which move the number — while the two things that do, fit and results, went unmanaged. You can’t improve what you never measured.

That’s the whole argument, closed. Retention isn’t a mystery and it isn’t hopeless — it’s been mismeasured. There’s one number underneath the dashboard that tells the truth about who actually stays, it sits at a readable height on your own retention curve, and it moves for exactly two reasons. The field has been flat for ten years because it was watching everything except that.

So the only question left is the one I’ve been building to for eight weeks: what’s the height of your plateau? Not the industry’s, not a benchmark’s — yours, read off your own customers’ behavior. It’s the number this entire series has been about, and you can now read it straight out of your own customer history. After a month of my argument, it’s time to see your own answer.