Last week I shared some data that surprised me: our churned accounts were outperforming every other "warm" segment we had, better conversion, better win rates, higher ACV, and almost nobody was working them.
The reaction I heard back more than anything else was some version of the same question. Okay, but how do you actually decide which churned accounts are worth going after?
That's the part most teams skip, and it's the part that actually matters. "Reach out to everyone who ever churned" isn't a strategy, it’s spray and pray.
You end up re-annoying accounts that left for good reason, and burning goodwill on accounts that were never coming back anyway.
When we ran this play, we built a filter before we built a single email.
The filter, not the funnel
The instinct most teams have is to jump straight to messaging. What do we say to win them back? Wrong question, or at least the wrong first question. The first question is which of these accounts are even recoverable.
Here's the filter we actually used:
Date of churn (recency). An account that left three months ago is in a different headspace than one that left two years ago. We used "churned since spring/summer 2020" to catch the COVID re-entry window.
Customer tenure before churn. A customer who stayed three years before leaving tells a different story than one who churned in month six. Longer tenure usually points to a situational reason, not a fit problem.
Previous ARR. We set a floor above our average ACV. Below that, the math on the effort doesn't work.
Churn reason, the dropdown. Filter out anything non-recoverable, a product limitation, a company that went out of business, and keep the rest: budget cuts, a champion who left, a competitor swap.
Churn reason, the actual notes. The dropdown tells you what happened. The CS and sales notes tell you why, and that's usually where your real messaging comes from.
Product usage before churn. High usage right before an account leaves usually signals a budget or political decision, not a product problem. Low usage means something else, and it goes in a different pile entirely.
Is the original champion still there? If yes, that's your way back in. If not, you're mapping a new stakeholder before you send anything.
Open opportunities. If sales already has a live conversation with the account, it's not yours to touch. Full stop.
That's the whole filter, and it took us from zero to six figures in pipeline and revenue in just six months.
The part that isn't a filter
Once the list was built, the messaging had to match the reason each account actually left. An account that churned over budget hears a different message than one that lost its champion. Send everyone the same generic "we miss you" email and you're right back to treating churn as one bucket instead of eight filtered ones.
We created five themes based on the churned reasons and sales notes, and worked with our PMM team to create messaging tracks for each. These tracks influenced the content we wrote for our assets, and outbound sequences.
Why this matters more than the tiering conversations
Most ABM advice will tell you to build a scoring model for this. Weight the signals, run it through a formula, maybe buy a tool that promises to rank your churned accounts for you.
Some of those things can work, but it’s of my opinion that you don't need a formula. You need clean CRM data and be able to answer eight honest questions about each account, and the discipline to leave the rest of the list alone.
I'm turning this into a full playbook, the CRM audit plus the execution framework for running the whole win-back motion end to end. That's out in two weeks.
If you've got a churned list sitting untouched in your CRM right now, is there a reason you haven’t focused on them, any specific blockers? Reply to the email, and let’s get you a plan in place.
What did you think about this week's newsletter?
