Your marketing attribution is lying to you (300% and counting)

A customer buys one thing. One order, €89, a Tuesday afternoon. By Wednesday morning three separate dashboards have each claimed it as their win.

Google Ads counts it as a conversion. Meta counts it as a conversion. Your email tool counts it too. One sale, €89, and the reports on your screen add up to €267 of “revenue we drove.” Three hundred percent of something that happened exactly once.

That’s marketing attribution in most accounts I open: mathematically impossible.

Why every channel claims the same sale

The mechanism is a bit embarrassing once you see it. Every platform reports conversions with its own pixel, on its own rules, inside a window that flatters it. None of them can see the others.

So the same buyer gets counted three times:

• Google Ads sees they clicked a Search ad six days ago. Its claim.
• Meta sees they viewed an Instagram ad on Sunday (a view, not even a click) and books it under view-through. Its claim.
• Your email tool sees they opened Tuesday’s newsletter an hour before checkout. Its claim.

Three tools, one receipt, three hands going up. Nobody’s lying on purpose. Each one is grading its own homework, and funnily enough they all give themselves an A.

The three lies stacked underneath

Last-click still runs the show. Most reports hand 100% of the credit to the final touch and nothing to the eleven touches before it. The Google ad that started the whole thing gets zero if the customer came back later through a branded search. One touch wins, the rest vanish.

Platform-reported conversions let each channel score itself. When Meta tells you it drove 41 sales, that’s Meta’s pixel, Meta’s attribution window, Meta’s definition of “drove.” You wouldn’t let a supplier write their own invoice and audit it too. That is exactly what a platform-reported conversion is.

Siloed pixels never compare notes. Google’s tag doesn’t know Meta’s tag exists. Neither knows what your email tool logged. There’s no referee in the room, so every overlap gets double- and triple-counted, and the totals balloon past anything real.

Stack those three and you get 300% attribution on a single €89 order. Nothing glitched. The system is built to do this.

What it costs to believe the dashboards

If you trust each report at face value, you fund every channel as if it earned the whole sale, and then you scale the channels that are best at claiming credit rather than the ones creating demand.

Retargeting is the classic offender. It swoops in at the bottom of the funnel and claims buyers who were already reaching for their wallet. On its own dashboard it looks like your hero. Turn it off for two weeks and total sales barely move, because it was harvesting, not hunting.

What honest marketing attribution actually needs

A “better” dashboard won’t fix this. Every platform will always over-claim, because over-claiming is how it justifies its own budget. The fix is one version of the truth they all have to answer to.

Four things make that possible:

• One source of truth, server-side. A server-side tracking setup records the event once, from your own server, before any platform gets to spin it. One clean log instead of three competing ones.
• Deduplication. When Google and Meta both reach for the same order, the server matches them on an event ID and counts the sale once. Overlaps collapse. The 300% falls back to 100.
• First-click to signed-contract tracing. Follow the customer from the very first ad, through every touch, to the order (or, for B2B, to the contract they sign weeks later). You stop rewarding the last click and start seeing which channel opened the door.
• Modeled, data-driven attribution instead of platform self-scoring. Spread the credit across the real journey with a model you control, not three suppliers each marking their own paper.

Done properly, the number in your own report finally matches the number in your bank.

Stop asking who gets the credit

“Which channel gets the credit for this sale?” has no honest answer, because the journey was shared and the platforms are biased witnesses.

Swap it for a colder question: “What happens to my total sales if I turn this channel off?”

That one you can actually test. Pause the channel and watch the real revenue in your bank, not the self-reported figure on its dashboard. If the total drops, the channel was pulling weight. If nothing moves, you just found budget that was buying credit for sales you’d have made anyway. Run a decent PPC management program this way and the winners show themselves fast, because switching them off visibly costs you.

The dashboards lie out of self-interest, not malice, so arguing with them gets you nowhere. Build one honest source they all have to reconcile against, and spend against that instead of three flattering fictions.

If you want the receipts to stop conflicting, start with the tracking. You can see how we scope and price that on the pricing page. The same order should be counted once, and you should only pay for what actually happened.