- Annual revenueA$1,200
- Bought on discount100%
- Return rate38%
- 12-month contribution-A$40
A win-back offer here pays a customer who already costs you, to keep costing you.
For Shopify operators
Your email tool knows who's about to leave. Not which ones are worth chasing.
A churn-risk score ranks customers by how likely they are to lapse. It says nothing about how much margin walks out with them. So your win-back budget chases the easiest customers to predict, not the ones worth saving.
The score you already have
You have a churn-risk score. It's a probability, not a P&L.
Klaviyo's predictive analytics already give you churn risk and an expected next order date. That part is genuinely useful. But they are built from order value and timing, not from contribution. The list ranks customers by how likely they are to lapse, and by how much they are predicted to spend.
Here is the gap. A high-risk customer who only ever bought on discount and returned half is not worth a win-back offer. A quietly-lapsing full-price regular is. Churn risk cannot tell them apart, because it has never seen your margin. So the win-back budget goes to whoever tops the risk list, not to whoever is worth keeping.
What it would change
Imagine your at-risk list ranked by the margin about to walk out.
Not the odds someone leaves. The profit that leaves with them. Order that list by contribution and the top of it looks completely different: the discount-addicts drop away, and the quiet, profitable regulars you were about to lose rise to the top, where your win-back budget can actually reach them in time.
- Spend win-back offers where they pay backSend the discount to the customers whose margin is worth the discount.
- Catch the profitable regulars in timeReach the full-price loyalists before a fixed flow ever fires.
- Let the money-losers goStop paying to keep customers who cost you on every order.
Why this happens
Churn risk ranks by likelihood. Value ranks by margin.
They are different lists. A customer can be high-risk and worth almost nothing, or lower-risk and worth a great deal. Rank your win-back by risk, or by predicted spend, and your best offers go to the wrong people. Weight it by contribution and the order flips.
High churn risk, low value: the discount-and-return buyer
Top of the risk list, but they only ever cost you. A win-back offer just pays them to keep losing you money.
Lower churn risk, high value: the quiet full-price regular
Further down the risk list, but this is the customer worth saving. Reach them and the margin stays.
Same two customers, ranked by the first bar and then by the second. Who deserves the win-back offer flips completely.
Check it in your own Klaviyo and Shopify data
Want to see it yourself? Here is how far the tools get you.
Klaviyo already predicts the churn. Shopify already holds the costs. The read you want lives in the join between them, which is exactly the join neither one does for you.
1Pull the at-risk listKlaviyo's churn-risk and expected-next-order predictions hand you the customers drifting away. That is your starting list.
2Join each to true marginFrom Shopify, bring in each customer's contribution: revenue minus COGS, shipping, fees, discounts and returns. Not their spend, their profit.
3Re-rank by margin at stakeSort the at-risk list by contribution, not risk. Now the top of it is the customers actually worth a win-back offer.
Where the tools stop
Klaviyo predicts the churn, but it has no COGS, shipping, fees or returns, so it cannot rank by margin. Shopify has the costs, but not the churn prediction. Joining the two, per customer, refreshed daily, is the part that falls to a spreadsheet, and it does not stay current for long.
Or the shortcut: Blufirescores every customer's drift and ranks the at-risk list by the margin at stake, straight to Klaviyo.
Worked through, in real numbers
Same risk score. Opposite worth.
Example 01Two customers, both 84% likely to churn
- Annual revenueA$980
- Bought on discount6%
- Return rate3%
- 12-month contributionA$430
Lower revenue, but real profit. This is the customer the offer should reach.
The risk score treats them identically. One is worth chasing hard; the other you should quietly let go. Only the margin tells you which is which.
Example 02Where the win-back budget lands
- Customers saved90
- Avg 12-month margin of a saveA$40 diluted by money-losers
- Margin recovered / monthA$3,600
A chunk of the discount lands on customers you would be better off without.
- Customers saved90
- Avg 12-month margin of a saveA$135 the ones worth keeping
- Margin recovered / monthA$12,150
Same effort, aimed at the customers whose margin actually pays the offer back.
Same win-back flow, same number saved. Choosing who by margin instead of risk recovers 3.4x the contribution. The budget did not change, the ranking did.
See the gap on your own numbers
Same win-back effort. Which ranking recovers more?
Enter how many customers you win back a month, and what an average save is worth over a year under each ranking. It shows you the margin you recover chasing churn risk versus chasing the margin at stake.
Customers you win back a monthHowever many your win-back flow actually reconverts
Chased by churn risk
A$3,600margin recovered a month
Chased by margin at stakeRecovers more
A$12,150margin recovered a month
Same win-back effort. Ranking your at-risk list by margin instead of churn risk recovers A$8,550 more a month , about 3.4x the return. You stop spending offers on customers you are better off losing.
A simple sketch. Blufire scores every customer's drift nightly and ranks the at-risk list by the margin at stake, straight to Klaviyo. Nothing you type here leaves your browser.
Where this stops being a spreadsheet
Blufire ranks the customers about to leave by the margin at stake.
It reads your store in contribution margin, revenue minus COGS, shipping, fees, discounts and refunds, reconciled to your ledger to the dollar. Then it watches every customer's own rhythm, flags the ones drifting away early, and hands you the at-risk list ranked by the profit at stake, ready to send from Klaviyo. The list your churn score was never able to build.
Keep going