Hey,

Quick one today.

I audited a home & furniture brand last week.

Average order around £500.

Their Performance Max campaigns reported a 6x ROAS, and everyone involved was understandably pleased. It’s the kind of number nobody questions.

Then we removed one thing from the report: conversions from searches containing the brand’s own name.

The 6x became a 2 to 3x. Nobody had touched the account.

We’d only changed what the report was allowed to count.

Nothing was technically fake.

PMax had simply been allowed to bid on branded searches, so every loyal customer who googled the brand to reorder got counted as a campaign win.

The platform was grading its own homework, and it gave itself an A.

That’s the part founders miss.

You don’t need broken tracking to get misled, because the default settings do it politely.

Meta ran the same trick in parallel.

The account sat on the default attribution setting, 7-day click plus 1-day view.

When we put the 7-day click column next to it, the default was reporting roughly DOUBLE the conversions.

Same ads, same month, same money. One column just takes credit for a lot of people who would have bought anyway.

Stack the two effects and a dashboard can show you 6x while the honest number sits near 2x.

Founders make real decisions on that gap.

They scale the wrong campaigns, keep the wrong agency, and skip fixes the business actually needs, because the screen says it’s working.

​
The 10-Minute Flattery Check

Three looks at your own account this week. You’ll be done before your coffee is:

1. Split brand from non-brand. In Google, pull the search terms report and separate anything containing your brand name. Judge PMax and Search on the non-brand number only. Exclude brand terms from PMax entirely if your setup allows it, and give brand its own cheap campaign. It’s the cleanest fix available.
   ​
2. Find your Meta flattery factor. Add the 7-day click & 1-day click column next to your default 7-day click + 1-day view. Divide one by the other. That ratio is how much your dashboard flatters you. Once you know it, you can’t unsee it, and every report after that reads differently.
   ​
3. Match the window to the buying cycle. A £500 considered purchase and a £30 impulse buy should never share settings. For high-AOV, longer-cycle brands I’ll typically set Google to a 30-day click window with a short view window, and sanity-check Meta scaling decisions against 7-day click trends plus cost per NEW customer.

The point of all three: platform numbers are inputs, and your contribution margin is the verdict.

I wrote about the other way ROAS lies, brand search and B2B contamination, here: Your 8x ROAS Is Lying to You​

None of this makes your ads worse, by the way.

Your ROAS gets smaller and your decisions get better.

That trade wins every single time.

Talk soon,
Andrej
​

PS: Reply “windows” and I’ll send you the attribution checklist we run on every new account, with the exact settings I’d use for your AOV and buying cycle.