The customer may see your name before they see your website, but that does not make every brand click worth buying

You are probably seeing the same pattern in your Google Ads account: branded search has the best conversion rate, the lowest apparent cost per lead and a report that makes it look indispensable. At the same time, Google answers more questions on the results page, AI summaries occupy attention and some customers never reach the website.

The useful answer is narrower than “yes” or “no.” Branded search advertising is still worth paying for when a competitor, reseller or comparison site can take the customer before your organic result does. It can also be worth paying when the ad sends people to a better destination than the homepage, such as a booking page, location page or current offer.

It is not automatically worth paying for when your business already owns the first organic result, the search has clear navigational intent and the ad is simply collecting credit for a visit that would have happened anyway. In that case, the impressive return in Google Ads may be a reporting artifact rather than new business.

AI summaries make the measurement problem harder, not the decision simpler

There is a real change in search behaviour. Pew Research examined 68,879 Google searches from March 2025. An AI summary appeared on 18% of them. When one appeared, users clicked a traditional result on 8% of visits, compared with 15% when no summary appeared. Only 1% clicked a link inside the AI summary itself. (Source: Pew Research Center, “Do people click on links in Google AI summaries?”) (pewresearch.org)

That does not mean every commercial search has become clickless. Pew’s sample included many informational searches, which are especially exposed to answer boxes. A person searching “what does a new roof cost” may accept a summary. A person searching “Smith Roofing emergency repair” is already trying to choose or contact a business.

Google’s own account is more optimistic. In August 2025, Google said total organic click volume was relatively stable year over year and that average click quality had increased. Google also says AI Overviews are driving more searches for the types of queries where they appear. (Source: Google Search, “AI in Search is driving more queries and higher quality clicks.”) (blog.google)

Those claims are not necessarily incompatible. Google can send stable total click volume while particular query types lose clicks. The practical conclusion for a small business is simple: stop treating all search traffic as one pool. Separate branded, non-branded, local and informational searches before deciding what to cut.

Keep branded search when it protects a sale. Pause or restrict it when it only harvests demand.

The strongest evidence says branded ads can be incremental, or almost completely unnecessary

Practitioners are reporting both outcomes, which is exactly why blanket advice fails.

A Silverback Strategies test for home-services provider CroppMetcalfe compared matched markets with branded ads running and paused. The agency reported that branded search was 100% incremental for that business. That is a useful warning for local operators: a brand campaign that looks cannibalized in a dashboard may still produce new customers when competitors are active or organic visibility is weak. (Source: Silverback Strategies, “Proof In the Revenue: Branded Search Drove Real Sales.”) (silverbackstrategies.com)

A different Search Engine Land analysis found the opposite. Of $36,129 spent on a branded campaign, only $3,945, or 10.9%, appeared to buy clicks that organic search could not have captured. The analysis classified 89.1% of spend as brand defense or cannibalization. After the campaign was paused, combined organic and direct revenue recaptured 65% of the paid-attributed revenue by week 13. (Source: Search Engine Land, “The $113,000 paid search test.”) (searchengineland.com)

LuckyRev reached a similar conclusion for an established ecommerce brand. Brand search represented more than 40% of Google Ads budget and reported 24.7x ROAS, but its media-mix model estimated much lower true ROI. After cutting brand-search spend by 75%, the company reported a 16% increase in net sales. That is a vendor case study, not a universal benchmark, but it shows why platform ROAS should not be mistaken for incremental profit. (Source: LuckyRev, “How MMM Revealed Brand Search Was Cannibalizing Revenue.”) (luckyrev.com)

AI does not remove the case for brand defence, but it changes what the ad must do

Google has begun placing ads in AI Overviews and testing them in AI Mode. Search, Shopping and eligible broad-match campaigns can appear in these experiences. Google says AI Overviews increased usage of Google by more than 10% for the types of queries that trigger them, based on its internal data from September 2024 through April 2025. (Source: Google Ads, “New ways AI in Search helps your business.”) (blog.google)

That creates opportunity, but it is not a reason to throw more money into a loose brand campaign. The ad still needs a job. It might reinforce trust when a searcher has seen your name in an AI answer. It might take the customer directly to “schedule service” rather than making them navigate a generic site. It might block a competitor on a high-value branded query.

It should not be used to make a weak website look present. If the landing page is slow, vague or missing the answer the searcher needs, paying to put it at the top only gets the disappointment there faster. This is the same reason to treat AI-generated marketing changes cautiously: An AI Should Not Change Your Website Until It Can Show Its Work. The ad, the search result and the website are one customer journey, not three separate reporting lines.

Seer Interactive’s study of 3,119 queries across 42 organisations found that, on its informational query set, paid CTR fell 78.4% year over year when an AI Overview appeared and the brand was not cited. The study also found higher CTR when a brand was cited. The figures are not a forecast for every local business, but they support a sensible shift: earn visibility in the answer, then make the paid result useful when the customer is ready to act. (Source: Seer Interactive, “AIO Impact on Google CTR: September 2025 Update.”) (seerinteractive.com)

Test the campaign against the sale you would have received anyway

1. Separate brand traffic from everything else

   Use a dedicated Search campaign for your name, common misspellings and brand-plus-service terms. Do not let Performance Max hide the result inside a blended total.

2. Record the auction before changing bids

   Check impression share, competitor presence, search terms, organic position and the landing page used for each important query.

3. Run a controlled pause

   Where the business can tolerate it, pause branded ads in matched locations or time periods while keeping other marketing as stable as possible.

4. Measure total conversions, not paid conversions

   Compare paid, organic, direct calls, form fills, booked jobs and revenue. A fall in paid conversions is not a failure if organic or direct conversions replace them.

5. Restore only the useful parts

   Keep brand defence where competitors appear, keep brand-plus-service terms that produce incremental jobs and cap or remove pure navigational terms that organic already captures.

For a small local business, the right answer is usually a narrow brand campaign, not a permanent tax

A small business rarely needs to defend every spelling of its name in every market all day. Start with the searches that carry commercial risk: your name plus “phone,” “near me,” a service, a location, a price-sensitive term or an urgent need. Write an ad that gives the customer a reason to choose the paid result, then send it to the page that completes that task.

If nobody else is bidding, your organic listing is strong and the customer is clearly looking for you, set a low cap or run a pause test. If a competitor is regularly visible, your organic result is buried under map packs and answer features, or the campaign drives calls that do not appear in website analytics, keep the campaign and judge it on total booked work.

Do not use the branded campaign’s conversion rate to justify itself. Brand searches are already high intent. The question is whether the ad changed the outcome. This is also why high-intent local search deserves separate treatment from social discovery. Do Not Move High-Intent Local Search Budget to Social Just Because Social Brings More Traffic makes the broader point: traffic volume is not the same thing as buying intent.

The decision is not whether Google deserves another dollar. It is what the dollar is buying

Google and AI summaries have made the search page more crowded and the click harder to interpret. They have not made branded search advertising universally useless. They have made lazy measurement more expensive.

Keep paying when the ad protects a vulnerable customer journey or creates a measurable extra sale. Restrict it when the campaign is collecting demand your organic result already owns. Test before making a permanent decision, and compare the whole business rather than the paid column in isolation.

For most small businesses, that means a modest, tightly matched brand campaign with clear exclusions and a scheduled incrementality check. It does not mean bidding on your name forever because the dashboard says the campaign has a high ROAS.