Channel ROAS consolidation: stop paying twice for the same click

You run Meta + Google Search + YouTube. Each platform reports its own ROAS. You add them up. The sum is wrong.

The over-counting mechanic

A prospect sees your Meta ad on Tuesday, Googles your brand on Thursday, then clicks a YouTube pre-roll on Friday before converting. In each platform's native reporting, that conversion is attributed in full — so you'd see:

• Meta: 1 conversion
• Google Search: 1 conversion
• YouTube: 1 conversion

Three ads. One conversion. In your monthly rollup that's triple-counted revenue and triple-inflated ROAS.

How bad is it typically

Across our last 20 audits, the median over-count was 22% of reported total conversions. The worst was 47% on an account running heavy brand-defense spend.

That means a reported 4.1× ROAS was actually sitting at 3.2× — and the brand-defense budget that looked like it was "working" was largely paying for intent the organic channels would have captured anyway.

What we do about it

First pass: de-duplicate using a deterministic identity key — email, device ID, or a first-party cookie we set before the first platform sees the user.

Second pass: run a channel-saturation model against your 12-month data. It tells you the incremental lift of each channel after the over-count is removed. Some channels turn out to be much smaller contributors than their native reports claim. Others turn out to be bigger — because they were capturing net-new demand that other channels later harvested.

Third pass: rebalance the spend. A ROAS goal of 3.5× on a net-of-overlap basis is far more defensible than a 4.5× goal on a natively-reported one.

The automation, not the dashboard

This runs as a weekly automation in our stack — no dashboard to log into. The output is a consolidated ROAS number, a spend-rebalance proposal, and a note on any channel whose saturation curve shifted.

If your CFO has ever asked "why do the platform reports add up to more than your bank says," this is the answer.