The purchase is the beginning of the next one
The weeks after an order are the highest-trust window a store ever gets with a customer. Most stores spend it sending a shipping notification and nothing else.
Post-purchase revenue is not about squeezing a bigger cart at checkout. It is about what happens in the ownership window: the accessory that makes the product better, the consumable it will need, the category the buyer is now statistically likely to enter. Retention runs this window on real cross-purchase data from your own orders, and Analyst keeps it honest by measuring repeat rate rather than send-level clicks.
Offers from co-purchase data, not guesses
What goes with what is an empirical question your order history already answers. The agent mines actual cross-purchase patterns per product and category, which routinely surface pairings a merchandiser would not have guessed.
Accessories and complements bought by previous buyers of the same item
Consumables and replacements timed to the actual usage cycle
Category progression — what first-time buyers of X buy next, and when
Offers suppressed where return history says the customer is not a fit
Timing is most of the outcome
The same offer performs completely differently at day two, day twenty, and day sixty. The agent times each offer to the ownership cycle of the product bought — setup accessories early, consumables at the replenishment point, category progression once satisfaction is established.
Measured on repeat rate, not clicks
A post-purchase programme that lifts clicks while flattening repeat rate is a failure. The agents measure cohort repeat behaviour against a holdout baseline, so you know the programme is creating orders rather than borrowing them from the future.
How it works
Mine co-purchase patterns
Analyst extracts what actually gets bought together and in what sequence from your order history.
Time the offers
Retention builds sequences timed to each product’s ownership cycle rather than a fixed calendar.
Measure against holdout
Repeat rate is compared to a baseline cohort, so the programme’s real contribution is visible.
What you get
Cross-sell offers derived from real co-purchase behaviour
Send timing matched to the ownership cycle of what was bought
Repeat rate measured against a holdout rather than assumed from clicks
Customers with poor-fit histories suppressed instead of pestered
Handled by these agents
Works with
Frequently asked questions
Is post-purchase upsell annoying to customers?
Badly targeted, yes. An accessory recommendation for something you just bought, timed when you need it, reads as service rather than sales. Fit and timing are the whole game, which is why guessing at them fails.
Should upsell happen at checkout or after?
Both exist, but they are different disciplines. Checkout upsell risks the primary conversion; post-purchase upsell risks nothing and can use ownership timing. The agent focuses on the second.
What about one-and-done product categories?
Then the play is category progression or referral rather than repurchase, and the data will say so. Forcing a replenishment pattern onto a durable goods catalogue is how programmes lose credibility.
How is success measured?
Cohort repeat rate and revenue against a holdout baseline over a meaningful window — not open rates, and not clicks, which are easy to lift while achieving nothing.
Keep reading
Increase average order value
Raise AOV with co-purchase data, threshold design, and offers that do not train discounting.
Run a post-purchase survey
One question at the right moment — attribution truth and objection intel from real buyers.
Subscription retention
Reduce subscriber churn with intervention before the cancel, not a survey after it.
Site speed optimisation
Find what actually slows real users down, fix it, and stop the regressions that undo it.
Put the team to work on your store
Connect your storefront, analytics, and email stack, set a goal, and let the agents run the work end to end. Start on the free plan with your own model key.