Module 05 · Profit
Bid to lifetime value, not first-order ROAS.
The ACOS that looks too expensive is often your best acquisition channel.
Most Amazon ad decisions optimise to the first sale — the ACOS on the order that ad generated. But a customer who reorders for a year is worth multiples of that first order. Judge ads on first-order ROAS alone and you'll quietly starve the campaigns acquiring your most loyal customers. And we don't guess that value from an ASIN's average — we measure it per customer, even on Amazon, where buyer identity is normally invisible.
The setup
First-order ACOS treats a one-time buyer and a year-long repeat customer as exactly the same. They're not. The product with a "bad" ACOS but a high repeat rate is quietly your best acquisition engine — and the rule that pauses it is costing you the customer, not saving you the click.
The proof
Count the real repeat, and the "expensive" ad target doubles
We don't infer value from an ASIN's average. We measure actual lifetime value per customer — resolving repeat purchases to the same buyer across orders, on Amazon and DTC alike. Pair real per-customer LTV with each ASIN's new-to-brand %, and the picture flips: products that acquire genuinely high-value customers justify far higher ad targets than first-order ACOS allows.
2–3×
the ad target the highest-LTV products justify once real per-customer value is counted. One turmeric SKU reached ~3.1× — its customers repeat at ~74% — while running on almost no ad spend. First-order ACOS called it "fine"; lifetime value says pour fuel on it.
How it surfaced: actual per-customer LTV joined to new-to-brand % — value the ad console never sees, because it stops counting at the first order.
And not just Amazon — the whole customer
Because we resolve the same customer across Amazon and your DTC store (the bridge behind Module 03), we can value them on their total cross-channel spend — not one channel's slice. An Amazon ad that wins a customer who then reorders on your Shopify store is worth more than Amazon-only LTV shows; the reverse holds when DTC ads feed Amazon. We credit the leak both ways and tune ad targets to the blended number.
Honest caveat: how much this moves the maths depends on your channel mix. For an Amazon-dominant brand it's a modest uplift. For a brand with a serious DTC store, blended LTV can change which campaigns are worth funding at all — which is exactly when it's worth knowing.
What it changes
→Push the high-LTV, high-NTB ASINs — fund the campaigns acquiring loyal new customers, even at a "scary" headline ACOS.
→Pull back the low-repeat ASINs — where the first order really is the whole story, an efficient ACOS is the right ceiling.
→Reframe the whole ads conversation — from cost-per-order to value-per-customer, with a number behind it.
How we see it
We resolve repeat purchases to the same customer — through the same Amazon-compliant identity layer behind Module 03 — for actual lifetime value per customer, not an ASIN-average guess. (Amazon masks buyer identity — so this is the hard part: we resolve it across all five marketplaces through proprietary, Amazon-compliant means; DTC is measured in full.) New-to-brand comes from the ads / AMC data. Combined per ASIN into an LTV × NTB map — the one number the console is missing: what the customer is worth after the first click.
The takeaway
Why this matters
The ad console optimises to the first click; lifetime value lives in an entirely different dataset. Only by reading them together do you bid to the customer instead of the click — and fund the acquisition your ACOS rules are quietly switching off. A third eye on what a customer is really worth.
Want to see which "expensive" campaigns are actually your best acquisition — and which "efficient" ones aren't?
Map my LTV × NTB →
The 2–3× targets and the turmeric example are from a real LTV × NTB analysis of one brand, anonymised. Amazon-side LTV uses identity-resolved purchase histories (proprietary, Amazon-compliant); DTC LTV is full-population.