There is a particular shape of account we see often enough to recognise from the first screenshot. Strong blended ROAS. Healthy-looking search performance. Growth that stopped about eighteen months ago and nobody can say why.
Almost always, branded search is doing the heavy lifting and the reporting is not separating it out.
What branded search actually does
Someone types your brand name into Google. They already know who you are, and they have already decided to look you up. Your branded campaign serves an ad above your own organic listing. They click the ad. They buy.
The campaign records the revenue and reports a ROAS that is frequently 10× or higher. Everyone is pleased.
But consider the counterfactual. Had the ad not been there, a meaningful share of those people would have clicked the organic result directly beneath it and converted anyway. You paid for a click you were going to get free.
That is not an argument for switching branded search off. There are real reasons to run it, and we will come to them. It is an argument for not counting it as acquisition, because it is not acquisition. It is a toll you pay on demand you already created.
The compounding problem
Here is why this matters beyond reporting hygiene.
Because branded search looks so efficient, it makes the blended account average look good. Because the average looks good, nobody digs into the components. Because nobody digs in, the genuinely weak part, non-brand acquisition, never gets fixed.
Meanwhile, branded search volume is a function of demand you built previously. As that earlier demand generation fades, branded volume declines, and the account's reported performance degrades for reasons that appear mysterious because the cause happened two quarters ago.
The account was never strong. It was living off stored demand and reporting the withdrawal as income.
How to check yours in ten minutes
Segment paid search revenue into branded and non-brand. Most accounts can do this by campaign, and if you cannot, that is your first finding.
Then answer three questions.
What share of paid revenue is branded? Above 50% is worth examining. Above 70% means your reported acquisition performance is largely fictional.
What is your non-brand ROAS on its own? This is your actual acquisition efficiency. For most brands it is a considerably less comfortable number than the blended figure.
Is branded volume growing or shrinking? Shrinking branded volume alongside flat total revenue means you are consuming demand faster than you are creating it.
What to do about it
The fix is uncomfortable in the short term and we always agree it with clients in advance, because it makes the headline number worse before it makes the business better.
Report branded and non-brand separately, permanently. Never blend them again. This single change fixes most of the decision-making downstream.
Judge acquisition only on non-brand. Set targets against it. Let branded search be measured as what it is, a defensive cost.
Move weight into mid and upper funnel. This is where reported ROAS falls, because the account stops claiming credit for demand it did not create. Contribution margin is the number to watch instead.
Test branded reduction carefully. In markets where competitors do not bid on your brand, you can often reduce branded spend substantially with minimal revenue loss. Test by geography rather than switching off nationally, and be aware that the moment a competitor starts bidding, the calculation changes.
What this looked like in practice
We took on a fashion brand where branded search carried over 80% of PPC revenue and growth had flatlined despite a healthy-looking account.
We ran profitability analysis by category first, then deliberately shifted weight out of branded search and into Performance Max and Shopping targeting mid and upper funnel demand.
Reported ROAS moved in ways that would have looked alarming without the prior conversation. Over the engagement, ecommerce revenue rose 32%, orders rose 54%, paid CPA fell 36%, and first-order contribution margin rose 21%.
The account did not get better at advertising. It stopped lying to itself about what its advertising was doing.
The honest summary
Branded search is worth running for most brands, defensively, at a controlled cost. It is not worth counting as growth, and any reporting that blends it into a single ROAS figure is making a weak account look strong at precisely the moment you need to know otherwise.
