For most established businesses, the most important search metric is no longer total organic traffic or average ranking. It is how many people search for the business by name, and what they find when they do. Generic, problem-led searches are increasingly answered on the results page itself. Searches for a specific business are among the searches AI features intercept least, and they measure something far harder for an algorithm update to take away: demand you created.
That is a stronger claim than it first sounds, so it is worth being precise about what the evidence supports and where it runs out.
The shiftWhat Has Changed in Generic Search
The research on AI features and clicks now points consistently in one direction. Pew Research Center tracked the browsing of 900 US adults in March 2025 and found that when an AI summary appeared, people clicked a traditional search result on 8% of visits, against 15% when no summary appeared, and clicked a link inside the summary itself on just 1% of visits. SparkToro’s 2026 analysis of US clickstream data estimated that 68% of Google searches between January and April 2026 ended without a click, against about 60% in 2024, though Rand Fishkin cautions that the two figures come from different data panels.
The pattern is not evenly distributed. In an Ahrefs study of 146 million desktop results pages from September 2025, AI Overviews appeared on 24.9% of non-branded results pages but only 13.1% of branded ones. An earlier Amsive study of roughly 700,000 keywords across ten websites found that only 4.79% of branded keywords triggered an AI Overview at all, and that where one did appear on a branded query, click-through rate rose by 18.68% on average, while non-branded keywords that triggered an AI Overview saw click-through rate fall by 19.98%.
The reasoning is intuitive once stated. An AI summary can answer “how long should a geyser last” without sending anyone anywhere. It cannot meaningfully answer “Smith Plumbing Johannesburg” on the person’s behalf, because the person has already decided where they want to go. Our article on what AI Overviews actually mean for your traffic sets out the broader effect on informational queries.
The argumentWhy Branded Demand Is Durable
Branded search is resilient for three reasons that go beyond AI features.
It reflects a decision already made. A person searching your name has heard of you, been referred, seen your work or used you before. A competitor’s content strategy struggles to intercept that intent, and a core update is far less able to redistribute it.
It accumulates as interaction history. In sworn testimony during the US government’s antitrust case against Google, Google’s Pandu Nayak described Navboost, a system that memorises past clicks on past queries over a 13-month window, sliced by locale and device, and described it as an important signal in ranking, though not the only one. A reasonable inference is that a business local people search for and click on is building that kind of interaction history in its own market, and that history cannot be bought directly.
It mirrors how Google’s raters are told to judge reputation. Google’s Search Quality Rater Guidelines require reputation research on every page quality task, directing raters to look for independent reviews and recommendations and to be sceptical of what websites say about themselves. Google’s own guidance on helpful content says that, of the E-E-A-T aspects, trust is the most important. A business people actively look for, and talk about elsewhere, is visibly trusted.
What Feeds Branded Search
If branded search is the outcome, the obvious question is what drives it. The most interesting evidence here comes from studies of AI visibility rather than classic rankings. Ahrefs analysed 75,000 brands and found that the number of branded web mentions correlated with how often a brand was mentioned in AI Overviews at 0.664, against 0.392 for branded search volume and just 0.218 for backlinks. These are Spearman correlations, Ahrefs is careful to say this is correlation, not causation, and its sample was filtered to domains with a Domain Rating above 40, which skews it toward established brands. But the direction is suggestive: being talked about off your own website appears to matter more than the link-building that dominated SEO for two decades.
For a South African business, the practical inputs are unglamorous and well known: reviews on Google and Hellopeter, mentions in industry publications and local media, a Google Business Profile that is complete and active, referral relationships, speaking and sponsorship, and work good enough that clients recommend it by name. Content plays its part when it is the kind people return to, share and cite. The opposite, as we have argued before, is why generic websites attract generic clients: nothing about them is memorable enough to search for.
First-party relationships matter too. A business with an email list, repeat customers and an audience it owns generates branded searches every time it communicates, which is one reason we argued for building first-party data after the cookie U-turn.
MeasurementHow to Measure It Properly
Google Search Console added a branded queries filter in November 2025 and extended it to all eligible websites in March 2026. It splits performance data into searches that include your business name, its variations, misspellings and brand-specific products, and everything else. Use it, with its limits in mind: Google classifies the queries with an AI-assisted system and says some may be misclassified, the filter is not available for websites with low impression volume, and it shows up to 16 months of history.
Once the split exists, read the two lines as different instruments. Non-branded clicks measure discovery: how well you win people who do not yet know you. Branded impressions measure demand: how many people already want you specifically. Branded clicks measure how much of that demand you actually capture. A business can have falling non-branded traffic and rising branded impressions at the same time, which can be a healthy business adapting to a changing results page, not a failing one.
Two common substitutes are weaker than they appear. Google Trends is sampled and normalised, and shows rare search terms as zero, so for most South African small and medium businesses it shows nothing at all. Direct traffic in Google Analytics is often treated as a brand measure, but it is really traffic Analytics cannot attribute to a clear source, which in practice also catches untagged links, documents and lost tracking. It is a noisy proxy at best.
Own the first page for your own name
Branded clicks are not automatically yours. Advanced Web Ranking’s third-quarter 2025 data showed desktop click-through rate for the first result on branded searches falling by 1.52 percentage points compared with the second quarter, while positions two to six gained a combined 8.71 points. Review platforms, directories, competitors’ ads and AI features all compete for clicks on your name. The defensive work is to make sure every result on your branded first page is one you are content with: your website, your Google Business Profile, your review profiles, your social and directory listings, all consistent and current.
“Branded search measures demand. It does not create it.”
This is the strongest objection and it is correct. Branded search is a lagging indicator. It reflects reputation built elsewhere, and a business that tries to “optimise branded search” directly, rather than doing the things that make people seek it out, is chasing a number rather than a result. It also says nothing about winning customers who have never heard of you, which is how every business grows beyond its existing network.
The evidence for non-branded visibility is not dead either. Seer Interactive ran 10,000 finance and SaaS questions through the GPT-4o API, the model behind ChatGPT at the time, and found that page-one Google rankings correlated with how often a brand was mentioned, at about 0.65. Generic search still feeds the awareness that later shows up as branded demand. And for a new or very small business, there may be no branded volume to measure at all.
Those points refine the position rather than overturn it. Non-branded search remains the discovery channel and still deserves investment. What has changed is how it should be judged. It is no longer reasonable to measure a business’s search health by total traffic, because a growing share of discovery now happens without a click. Branded demand is where that invisible discovery eventually becomes visible.
Stop reporting a single organic traffic number. Split it. Treat non-branded search as your discovery investment, judged over quarters and expected to deliver fewer clicks per impression than it once did. Treat branded impressions as the health score of the business itself, arguably the truest available measure of whether your reputation is growing. Then put real effort into the things that move it: work worth recommending, reviews, mentions, an audience you own, and a first page for your own name that you control. Topical authority still builds discovery. Branded demand is the proof it worked.
Frequently Asked Questions
What is branded search?
Branded search refers to searches that include a business’s name, its variations or misspellings, or its specific products, for example “Wired Web Services maintenance”. Non-branded searches describe a need without naming a business, such as “WordPress maintenance Johannesburg”.
How do I see branded search traffic in Google Search Console?
Use the branded queries filter in the Performance report, which Google introduced in November 2025 and extended to all eligible websites in March 2026. It is not available for websites with low impression volume, and Google notes that its AI-assisted classification may misclassify some queries. Smaller websites can approximate it by filtering queries that contain the business name.
Do AI Overviews appear on branded searches?
Less often than on generic searches. An Ahrefs study of September 2025 results pages found AI Overviews on 13.1% of branded results pages against 24.9% of non-branded ones, and an Amsive study of ten websites found that branded click-through rate rose on average where an AI Overview did appear.
How can a small business increase branded searches?
By giving people reasons to look for it by name: work worth recommending, steady Google and Hellopeter reviews, mentions in local and industry media, an active Google Business Profile, regular contact with existing customers, and content people return to and share.
Is branded search a Google ranking factor?
Google has not said so directly. Sworn testimony in the US antitrust case described Google using past click data in ranking, and Google’s rater guidelines emphasise reputation, but there is no confirmed “brand” ranking factor. Branded search is best treated as a measure of reputation and demand, not a lever.
Give People a Reason to Search for You
For businesses on our maintenance plans, our SEO and AI Visibility service tracks rankings on Google and Bing alongside how AI assistants mention the business. If your website itself is not yet giving people a reason to remember you, our website design services team can help, and you can get a quotation directly.
Or reach out through our contact page.
