A pattern I see again and again in digital diligence: paid-media ROAS that looks great on the headline number and falls apart under five minutes of structural questioning. The mechanism is almost always the same. A meaningful proportion of the spend is going against the brand's own keywords, where the customer was going to find the brand anyway.
This is a known issue. It remains one of the single largest sources of overstatement in the digital businesses I diligence.
The reason it persists isn't that buyers are gullible. It's that brand bidding looks like good performance until you ask the second question.
What's actually happening
A user types the brand name into Google. The brand has bid on its own term. The user clicks the paid ad — sometimes because it's at the top of the page, sometimes because they didn't notice it was an ad. The conversion happens. The paid platform attributes the conversion to the paid click.
ROAS looks excellent on that traffic. CPCs are low because the brand has a high quality score on its own terms. Conversion rates are high because the user already knew what they wanted. The cost-per-acquisition is a fraction of non-brand paid CPA. The unit economics on this slice of spend are, by every measure that matters to a paid media dashboard, beautiful.
The problem: a meaningful share of those users were going to convert anyway. They typed the brand name. They were arriving at the brand's site. The conversion that's now attributed to paid would have happened with no paid spend at all — or at much lower paid spend, or with paid spend redirected somewhere it could actually generate incremental demand.
The blended ROAS number that includes this brand spend overstates the asset's paid-media efficiency. Sometimes by a lot.
How to identify it
The diagnostic is straightforward, but it has to be done. Three steps.
Pull the search-term report. Filter to the brand's name and obvious variants. Sum the spend, the clicks, the conversions, the conversion value. That's the brand-bidding slice.
Calculate the share of paid spend it represents. In a healthy paid-media operation, brand spend is a small minority of total spend — typically under 15%, sometimes much less. In an operation where the headline ROAS is being inflated, brand spend is often 30–50% of total. That's the smoking gun.
Calculate the share of paid conversions it represents. Brand-keyword conversions usually punch above their weight on conversion rate, so the share of conversions is often even higher than the share of spend. This is the part that flatters the ROAS.
Now strip the brand spend and brand conversions out and recalculate ROAS on the non-brand-only slice. That's the asset's real paid-media efficiency. It's almost always materially worse than the headline number.
Why brand bidding persists despite the issue being known
If everyone knows about this, why does it keep happening? Three reasons.
It's not zero-value. Some brand bidding is genuinely defensive — keeping competitors off the brand's own SERP, capturing users who were going to find the site but might have got distracted by an ad above. The defensive case has merit. The problem is that "defensive" gets used to justify a much larger share of spend than the genuinely defensive case requires.
It looks great in the dashboard. The agency or in-house team running the account is incentivised to show good numbers. Brand spend produces the best numbers in the account. A team that pulls back brand spend is, on the dashboard, performing worse — even if the business's actual unit economics have improved.
It pads the line. If the in-house marketing team needs to demonstrate paid-media performance to leadership, brand spend is the easy way to do it. The headline ROAS goes up. The pipeline stays the same.
For a buyer in diligence, the question isn't whether brand bidding is happening. It's how much of it is happening, what proportion of headline performance it's contributing, and what the asset's paid efficiency looks like once you've stripped it out.
What this means for valuation
The correction is straightforward but the implications are not always priced into the deal.
If a buyer is paying a multiple on the asset's marketing-driven revenue, the multiple should be applied to the genuinely marketing-driven slice — not the slice that includes brand demand the asset would have captured anyway. In practice this means: if the asset is reporting blended ROAS of 4.0 and the brand-stripped ROAS is 2.5, the asset's paid-media performance is much weaker than it looks. The hold case for paid scaling post-deal is correspondingly weaker. The 100-day plan needs different inputs.
I've seen deals where the brand-stripped paid efficiency was so far below the blended number (one of the five red flags I find on most engagements) that the buyer's investment thesis didn't survive the correction. Better to find this in week two of diligence than month two of ownership.
What sellers should know
For sellers reading this — and I know some of you do — a quick note. The diligence framework on this is getting more sophisticated, not less. Five years ago, a seller could plausibly present headline ROAS and not be challenged on the brand-bidding share. That's not the case now. Sophisticated buyers run this analysis as standard. The seller who's prepared a defensible breakdown of brand vs non-brand spend, with an honest commentary on the strategic rationale for each, will fare materially better in negotiation than the seller who waits to be asked.
The asymmetry of preparation is, again, doing real work in deals.
A practical recommendation
If you're a buyer running diligence, get the search-term report (and run the broader 30-minute paid audit framework on the rest of the account). If you're a seller preparing for sale, run this analysis on yourself before the buyer does. If you're an in-house marketer or an agency running paid for a brand that might be sold in the next 24 months, start producing the brand-stripped ROAS number alongside the headline number now, in the regular reporting. It will hurt for a quarter or two. After that, it'll be the number you wish you'd been reporting all along.
The brand-keyword trap is one of the most common ways paid-media diligence goes sideways. It's also one of the easier ones to spot once you know to look. Buyers, ask the question. Sellers, answer it before you're asked.