Home/Research

Paid search · natural experiment

Brand search ads do not add traffic. They replace it.

An account suspension gave us a natural experiment across three online stores. Organic click-through collapsed while brand ads ran and recovered above baseline within a week of them stopping — on two of three. The third is why this is a test, not a rule.

Published 2026-08-04

What we were looking at

A brand search campaign bids on your own name. The dashboard reports it as one of the best campaigns you have: high click-through, low cost per acquisition, excellent return on ad spend. That is exactly what you would expect if the people clicking were going to arrive anyway.

The question is not whether brand ads convert. It is whether they are incremental — whether total traffic goes up when they run, or whether paid clicks simply replace free ones.

The natural experiment

One store's Google Ads account was suspended, which stopped its brand campaign overnight without us choosing the timing. That removes the usual objection to a before/after: nobody picked the moment.

While the brand ads were running, organic click-through on the store's own brand queries fell from 57.5% to 35.3% to 22.0% across three weeks. Impressions were stable and average position held at 1.1 — so this was not a ranking loss and not a demand loss. The clicks were going somewhere else: to the ad above the organic result.

In the week after the ads stopped — ten paid clicks in total — organic click-through returned to 74.6%, above where it had started. Full recovery in one week. No competitor took the traffic in the meantime, which is the argument brand campaigns are usually defended with.

What it looked like on two more stores

The same signature showed on two others: brand impressions flat, position stable, organic click-through falling as brand spend rose. One went from 79.2% to 19.6%, its organic clicks from 1,749 to 487 a week. The other from 77.2% to 33.8%.

We paused the brand campaign on both and watched revenue rather than clicks, because clicks are not the thing being bought.

Where it did not hold

On one of the two, organic revenue rose and covered the loss almost exactly — total search revenue stayed flat while the brand spend went to zero. Two weeks later organic was still 38% above its pre-pause baseline. That one stayed paused.

On the other, organic revenue fell and stayed down. Total search revenue dropped by about a third over two weeks. The ads there were incremental — they were bringing people who did not otherwise arrive. We restarted the campaign.

That is the finding, and it is the reason this is written as a method rather than a rule. One of the two stores we were confident about was the opposite of what we expected.

A counterexample worth knowing

A fourth store's campaign looked like the same waste by name — it had “Brand” in the title — and was not. Only part of its spend was on the store's own name; the rest was on generic terms and on a competitor's product name. Impression share on its brand terms was 25.5%, with 63% of the loss to rank rather than budget — meaning somebody else was bidding and winning. And its organic brand click-through had never been high to begin with, so there were no free clicks to lose.

Brand ads are waste only where you already own the click and nobody is contesting it. Do not judge a campaign by its name.

How to run the test yourself

1. In Search Console, split queries into brand and non-brand and pull them weekly.

2. Look for the signature: impressions stable or rising, average position stable, organic click-through falling. That combination is cannibalisation. Falling impressions would be a demand problem; a falling position would be a ranking problem.

3. Pause the brand campaign on one account. Not all of them.

4. Measure revenue, paid plus organic together, for at least two weeks. One week is a spike, not a recovery.

5. If organic does not take the traffic back, the ads were incremental. Restart them and say so.

Limitations

Four stores, one market, one language, cash-on-delivery heavy. The suspension was a genuine natural experiment; the three deliberate pauses were not blind and were not randomised. Revenue figures are our own stores' — no client data is involved, and no budgets, margins or spend figures are published here.

The percentages above are click-through rates and impression shares as reported by Google Search Console and Google Ads for the weeks described.

Questions and answers

Are brand search ads worth paying for?

Brand search ads are worth paying for only if they are incremental, meaning total traffic rises when they run instead of paid clicks replacing free organic ones. High click-through and low cost per acquisition do not prove this, because people searching your own name may arrive anyway. The reliable test is to pause brand ads on one account and measure paid plus organic revenue together.

How do I know if brand ads are cannibalising organic traffic?

Brand ads are cannibalising organic traffic when brand impressions stay stable or rise, average position stays stable, and organic click-through on brand queries falls as brand spend rises. Falling impressions would point to a demand problem instead, and a falling position would point to a ranking problem. Pull brand and non-brand queries from Search Console weekly to see the pattern.

What happens to organic clicks when brand search ads stop?

When brand search ads stop, organic click-through on brand queries can recover quickly. In one natural experiment caused by an account suspension, organic click-through returned above where it had started within one week, and no competitor took the traffic. That result did not hold everywhere, though: on another store, revenue fell after the pause, so the outcome has to be measured, not assumed.

How long should a brand ads pause test run?

A brand ads pause test should run for at least two weeks, measuring revenue from paid and organic search together rather than clicks. One week shows a spike, not a recovery. Pause the brand campaign on one account rather than all of them, and if organic search does not take the traffic back, the ads were incremental and should be restarted.

When are brand search ads actually needed?

Brand search ads are needed when they bring people who would not otherwise arrive, which shows up as total search revenue falling after a pause. Brand ads also matter where someone else is bidding on the brand terms and winning, or where organic brand click-through was never high, so there are no free clicks to lose. Brand ads are waste only where the click is already owned and uncontested.

Why is a campaign named brand not always a brand campaign?

A campaign named brand is not always a brand campaign because its spend can go partly to generic terms or to a competitor's product name rather than the store's own name. In one counterexample, low impression share on brand terms lost mostly to rank showed that another advertiser was bidding and winning. Judge a campaign by where its spend goes, not by its title.

All research