What is off-site ROAS and why does platform reporting miss it?

Your Meta account says the campaign lost money. Amazon, Walmart, and Target say it paid for itself three times over. Both are looking at the same ads. Only one of them is looking at the sales.

What is off-site ROAS?

Off-site ROAS is the return on ad spend for orders that happen somewhere other than the domain where the ad's tracking pixel lives. Spend $10,000 on Meta and drive $40,000 in revenue at Amazon, and your off-site ROAS is 4.0. Meta's dashboard, meanwhile, reports a fraction of that or nothing at all.

The metric matters because most consumer brands don't own the checkout for most of their orders. They sell through retailers. The ad platform sees the click, the retailer sees the purchase, and no report joins the two sides.

Why does platform-reported ROAS miss it?

Meta's pixel fires on your domain. When a shopper clicks a Meta ad and lands on your product page, the pixel tracks their session. If they buy on your site, Meta records a purchase event and attributes revenue to the ad. Reported ROAS looks correct.

The moment that shopper leaves for Amazon (through a Buy on Amazon button, a link in your bio, or a search they run after seeing the ad), the pixel goes blind. Amazon does not report sales back to your Meta ad account. Neither does Walmart, and neither does Target. The purchase happens, the revenue lands in your bank, and the ad that caused it shows a zero.

For DTC-only brands, this gap is small. For brands whose sales split across a website and one or more retailers, the gap is the majority of the business.

A worked example: the marketplace brand

Take a mid-sized home-goods seller spending $50,000 a month on Meta ads. The catalog lives on the brand's Shopify store, Amazon, Walmart, and Target. Here is where the traffic goes in a typical month:

True ROAS across all destinations: $200,000 / $50,000 = 4.0.

Platform-reported ROAS, based only on the on-site 18%: $36,000 / $50,000 = 0.72.

A media buyer in Ads Manager sees a campaign losing 28 cents on every dollar. The finance team looking at retail sell-through sees a campaign returning $4 for every $1. Both numbers are computed correctly. They measure different universes.

The breakeven ROAS for this brand is 1.6. On the visible number, every ad should be paused today. On the true number, every ad should be scaled tomorrow. Whichever way the team goes is a coin flip driven by which report they happen to trust.

What the missing signal does to campaign delivery

The reporting gap is the visible problem. The invisible one is worse.

Meta's delivery system chooses who sees your ad based on conversion signal. When it sees a purchase event tied to an ad view, it learns that people like that viewer buy, and it goes hunting for more like them. When it sees only clicks and no purchases, it learns that people like that viewer click and then vanish. So it goes hunting for cheap clickers.

Over weeks, this compounds. Audiences drift toward window shoppers. Lookalikes get built from seeds of people who bounced. CPMs stay low, CPCs stay low, and reported conversions stay near zero, because the algorithm is doing exactly what the data told it to do. The feedback loop broke somewhere between the click and the retailer's checkout, and Meta has no way to know it happened.

So off-site ROAS is more than a reporting metric. The platform needs the signal to do its job.

The three ways brands try to close the gap

Brands aware of the problem usually reach for one of three approaches:

1. UTM tagging. Add campaign parameters to retailer links. This tells you a click happened. It does not tell the retailer to report a sale, and it feeds Meta no purchase signal. Useful for post-hoc analysis, useless for delivery.
2. Retail media networks. Move budget from Meta into Amazon Ads or Walmart Connect. Attribution works cleanly inside the walled garden, but you pay retailer CPCs (often 3 to 5 times Meta's), and you can't run cross-retailer campaigns from a single audience.
3. Closed-loop measurement. Capture the click-through-to-retailer as a server-side event and send it back to the ad platform's Conversions API. Meta then treats retail-bound clickers as a real conversion signal and steers delivery toward buyers rather than browsers.

Only the third approach fixes both the reporting number and the delivery problem at once. The first tells you what you already suspected. The second is a budget shift, not a measurement fix.

Where to start

• Website: www.pixamp.io. Send Meta traffic to retailers and get the buyer-intent signals back. First 1,000 clicks free, no card required.
• How it works: www.pixamp.io/#how-it-works. Three steps: 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.

Off-site ROAS is the number your CFO already trusts and your ad platform can't see. That gap decides whether you scale a winner or pause it by mistake.