The friction wasn't working

To stop paying us, a customer went through seven screens and then picked up a phone. Every screen was a save attempt, and every save attempt had been fought for by somebody with a retention number attached to it.

I was angry about it on behalf of our customers, and I’ll say that plainly. Making leaving hard is not a growth strategy, it’s a way of taxing people who have already decided.

When the deadline landed, marketing panicked, and they were right to. They own cancel mechanics, they are measured on retention, and someone had just told them to delete the machinery that produces their number. In a week.

It would have been easy to treat them as the obstacle. That would also have been wrong. The useful move was to separate two things that had been fused together: the offer, and the friction. Customers leaving QuickBooks genuinely might want a lower tier, a pause, or help with the thing that made them want to go. That support is worth offering. It just isn’t worth trapping someone to deliver.

So we kept the support and the incentives, gave them to people plainly, and made the cancellation itself take one screen. Nobody had to lose an argument. The offer and the obstacle were never the same thing, they had just been shipped together for so long that the team had stopped seeing the seam.

That gap is the whole case study. Marketing was defending seven screens against a twenty percent collapse in retention. We removed five screens and a phone call, made leaving genuinely easy, and roughly one percent more people left.

Which means the friction was never retaining anyone worth retaining. It was collecting a small tax in exasperation from people who were leaving regardless. Seven screens bought a percentage point and cost us the goodwill of every single person who went through them, including the ones who stayed.

I held the pen on what the screens said. Legal, marketing, product, and engineering moved just as fast, and a week is not a solo achievement.