An agency playbook for clients who sell on Amazon, not their site

The account looks broken on paper. You're spending, the client's Amazon sales are climbing, and Ads Manager shows conversions near zero. The campaign works. The report says it doesn't.

Why do marketplace clients break your reporting?

When a client's checkout lives on Amazon or Walmart, the Meta pixel never sees the sale. The click lands on a retailer domain you don't own, the purchase happens there, and Amazon does not report that purchase back to the ad account. Meta records a click, then nothing.

That one line changes how you run the account. Delivery learns from the conversions you feed it, so feed it clicks and it finds clickers. With retailer-bound traffic, clicks are all it ever gets, and the budget goes toward cheap curiosity instead of buyers.

For an agency, the damage runs both ways. You make worse buying decisions on blind data, and you can't prove the value you're already creating. Either one can cost you the account.

How do you set ROAS expectations on day one?

Set them before the first dollar spends. A client who sells on Amazon cannot expect platform-reported ROAS to match reality, because the platform only counts what happens on its own domain. Say that plainly in the kickoff.

Give them the arithmetic instead of an apology. If 7 of every 10 conversions happen at the retailer, Meta's report is missing 70% of the revenue the campaign drove. The Ads Manager number isn't conservative, it's wrong, and decisions built on it drift toward the wrong audiences.

Then reset the benchmark. The comparison that matters is client revenue against ad spend, pulled from Amazon or Walmart's own numbers, not the ROAS column in Ads Manager. Agree on that source of truth in writing. The same discipline that defines a breakeven ROAS target applies here, except the revenue side comes from the retailer, not the pixel.

How do you structure campaigns around buyer-intent events?

Meta's delivery gets smart only when you feed it a buyer signal rather than a click. That means routing traffic through a page you control, capturing the moment a shopper heads to the retailer, and sending that event back through the Conversions API.

The event you feed matters more than the volume. A buyer-intent signal tells the algorithm which audiences produce people who actually walk to a retail checkout, so lookalikes seed from buyers instead of window shoppers. Server-side delivery also means the signal survives iOS privacy settings and ad blockers.

Structure the account around that. One clean intent event per retailer, mapped to your existing pixel, so campaigns keep running while delivery rebuilds. Within about 48 hours of clean signal, Meta starts steering toward retail buyers, and your bidding decisions rest on real data.

How do you report retail impact clients trust?

Clients trust a report they can reconcile against their own bank. Build every deck to tie Meta spend to retailer revenue, and name where each number comes from. Never present platform ROAS as the headline for a marketplace account. It undersells you and teaches the client to distrust the tool.

A steady cadence beats a beautiful one. Pick reporting intervals that match how fast each number stabilizes, and hold to them. Retail sales data lags the click, so daily ROAS swings mean nothing. Weekly and monthly views tell the real story.

The two-week view is where you win renewals. It's the first cadence where retailer revenue is stable enough to divide by spend and show a number the client believes. Anchor the relationship on that number.

A 30-day onboarding plan for a new marketplace client

Run every new marketplace account through the same sequence so expectations and data land together.

1. Kickoff: agree that retailer revenue, not platform ROAS, is the scoreboard, and get the retailer reporting access in writing.
2. Wiring: connect Meta Business Manager with one OAuth click, add a retailer button to the client's product page, and confirm intent events reach the existing pixel.
3. Baseline week: let delivery collect clean buyer-intent signal without touching budget, and record where reported ROAS starts.
4. First reallocation: at day 7, shift budget toward ad sets producing the cheapest intent events.
5. First reconciliation: at day 14, put Meta spend next to retailer revenue in one view and set the blended ROAS benchmark.
6. Monthly review: at day 30, report reorder trend and blended ROAS, then decide scale.

This is also the cleaner answer to the deeper account problem, which is what happens to Meta's algorithm when you sell on retailers and don't track purchases. Feed it buyers and the account stops fighting you.

Where to start

• Website: www.pixamp.io. What Pixamp does, pricing, and the FAQ. First 1,000 clicks free, no card required.
• How it works: www.pixamp.io/#how-it-works. The three-step setup: connect Meta Business Manager, add a retailer button, launch. Live in under an hour.
• Book a demo: www.pixamp.io/#contact. A 20-minute walkthrough on a real retailer page, with the founding team.

If your client's checkout lives on Amazon and your reports still end at the click, you're leaving both performance and proof on the table. Closing that gap is an afternoon of setup per account.