Search campaigns took 41% of Google Ads clicks in August 2026 across 31 accounts managed by Brainlabs, down from 62% in January 2024, while their share of spend fell from 86% to 71%, according to a post the agency published on October 7, 2026. The piece, titled "Rethinking the Value of a Search Click" and written by Alessio Messaggi, argues that a growing share of buyers now compare products inside AI assistants before they ever reach an ad, and that paid search has no view of that stage. It pairs the click data with a 54-client organic traffic study and early ChatGPT ad pricing of £15 to £60 per thousand impressions.

In Short

A big advertising agency called Brainlabs looked at its own clients' numbers and found that people are visiting websites less often from Google's normal search results, and that Search ads now bring in a smaller share of the clicks its clients pay for. The agency thinks this is because many shoppers now ask tools like ChatGPT to compare options first, so by the time they click, they have mostly made up their mind. Its point is that counting clicks says less than it used to, and the value of each click, plus how to measure the conversations that happen before it, matters more.

What Brainlabs published

The post sits on the Brainlabs blog under its Paid Media resources and is dated October 7, 2026, according to the agency's blog index. It is opinion backed by the agency's own data, not a peer-reviewed study, and Brainlabs sells the services the post discusses. The site footer lists AI Visibility, Paid Search and SEO among its offerings and describes Brainlabs as "the most agentic media agency." Every figure below is attributed to Brainlabs unless another source is named.

The core claim is a change in where decisions happen. According to Brainlabs, paid search has long been "planned, bid on and reported around the click," on the assumption that a visitor arrives undecided. "That assumption has been wearing out for some time now," the post states, "and an ever growing share of the people who click have already made up their minds."

To illustrate the point, the post uses a family buying travel insurance for a ski trip. In the older pattern, the buyer types a few words into Google and works through five or six tabs of comparison sites and policy pages. In the newer one, the buyer asks ChatGPT to compare three policies for a family of four skiing in Austria in February, with medical cover for a parent who has a heart condition. "The reading, comparing and eliminating all happen inside the answer," according to Brainlabs. "The click is now one of the last steps in the decision."

The numbers behind the argument

Organic sessions across 54 clients

The organic figures come from an earlier Brainlabs analysis, published on September 14, 2026 under the title "Organic traffic is falling, but AI isn't replacing it," and reproduced as a table in the October post. Brainlabs compared the eight months after Google's AI Overviews became common with the eight months before. Neither post gives the exact calendar dates of the two windows.

Organic search sessions fell from 140,132,402 to 125,439,245, a decline of 10.5%. Organic key events fell 15%, from 47,386,398 to 40,286,450. New users from organic search dropped 12.6% and returning users 6%.

AI referral traffic moved the other way. Sessions rose 163.5%, from 502,396 to 1,323,650, and key events rose 335%, from 107,562 to 468,097. Yet the base was small. On a PPC Land calculation from the published table, organic search lost about 14.69 million sessions, while AI platforms added about 821,000, roughly 5.6% of the shortfall. That matches the post's statement that AI visits "replaced only around 5% of what was lost." Even after the increase, AI referrals accounted for about 1% of the combined organic and AI sessions in the post-Overviews period.

Brainlabs attributes the decline to how Overviews are deployed. According to the post, the agency's SEO team reads the data as Overviews triggering "overwhelmingly on informational searches" while "starting to reach commercial ones, so the research stage is increasingly finished inside the answer."

Search's falling share of the clicks advertisers buy

The second data set covers 31 Google Ads accounts, which Brainlabs describes as ranging "from high street retailers to insurers and charities." Two charts, labelled "Spend Media Mix L32M" and "Clicks Media Mix L32M," track the split between Search campaigns, Performance Max with Shopping, and Demand Gen with video from January 2024 to August 2026.

Search's share of spend fell from 86% to 71% over that period. Its share of clicks fell further, from 62% to 41%. Neither line moved smoothly. Read from the chart labels, Search's spend share dipped to 64% in December 2024 and 56% in November 2025, recovering to 80% and 74% in the following months, which suggests a seasonal shift toward Shopping and Performance Max in each fourth quarter. Its click share touched 31% in December 2024 and 33% in May 2026. In spring 2026, the Demand Gen and video band widens sharply on the clicks chart, approaching a quarter of clicks around May before narrowing again. A fourth series, labelled "Search Click N Deviation," is plotted on the clicks chart without explanation. "Search still takes most of the budget, but it now delivers well under half of the clicks," according to Brainlabs.

This is a portfolio measure, not an industry one. The 31 accounts are Brainlabs clients, and the shift toward Performance Max, Shopping and Demand Gen reflects buying decisions the agency itself made for them. The post links the trend to its thesis that keyword-based buying captures less of the decision, but it does not separate that effect from changes in Google's campaign products or from the agency's own allocation choices.

Conversion rates and the 1.5 times claim

The clicks that do arrive from AI platforms convert better, according to Brainlabs. In the post-Overviews period, the key event rate for AI referral traffic was 13.4%, against 10.4% for organic search, a gap of about 1.3 times. Before Overviews became common, the order was reversed: organic converted at 10% and AI referrals at 8.2%.

The post then states that "by April 2026 the gap had widened to roughly 1.5 times." A second table, with monthly data from January 2025 to April 2026, shows organic sessions falling from 18,737,103 to 14,072,146 over that span, a drop of about 25%, while AI referral sessions rose from 27,786 to 196,866. Measured against organic volumes, AI referral sessions grew from 0.15% to 1.40% of organic sessions, and AI referral key events from 0.1% to 1.82% of organic key events.

That 1.5 times figure cannot be reproduced from the published tables. Dividing key events by sessions for April 2026 gives about 0.45 for AI referrals and 0.35 for organic, a ratio near 1.3. The comparison is imperfect, because GA4's session key event rate counts sessions with at least one key event, as PPC Land reported when Google added the metric in June 2024, while the tables count total key events. Brainlabs does not say how the April figure was derived.

The monthly table shows a further pattern the post does not discuss. In January 2025, AI referral visits produced about 0.13 key events per session against 0.34 for organic, well below parity. The ratio crossed 1 only around January 2026. The conversion advantage that the post's argument rests on is, in this data, a recent development.

Fewer clicks, more conversions

The post's central arithmetic is an illustration built on those rates. A campaign that brings 1,000 clicks at a 10.4% key event rate delivers 104 conversions. A second that brings 800 clicks at 1.5 times that rate, or 15.6%, delivers about 125. "A report that counts clicks calls the second campaign the weaker one," according to Brainlabs, "while the business has gained around 20 more conversions from 200 fewer clicks."

The example uses the 1.5 multiplier that the tables do not confirm. At the 1.3 ratio visible in the April 2026 data, 800 clicks would produce about 108 conversions, still more than 104, but with a margin of four rather than 21.

On bidding, the post describes passing accurate first-party conversion value back to the platforms as the established response, so that automated bidding can compete for the clicks that earn more. "A higher value per click widens the inventory a business can afford to win," according to Brainlabs.

ChatGPT ads and their pricing

The post presents ChatGPT's ads as "an early way to pay to appear inside the conversation." According to Brainlabs, targeting is contextual to the prompt, and "with no conversion optimisation yet," buyers pay for impressions and clicks and work out their value separately. Early CPMs, it says, have ranged from about £15 for niche placements to £60 for competitive ones, averaging close to £20 to £25.

The pricing roughly tracks what PPC Land has reported in dollars. ChatGPT ad CPMs launched at $60 on February 9, 2026 and fell to as low as $25 by April, with Ad Age citing figures as low as $15. Brainlabs quotes the same numerals in pounds and gives no dates, markets or sample for its range. Because the agency works across the UK and US, the post does not make clear whether these are UK prices, converted US prices or a mix.

The statement that there is no conversion optimisation also differs from OpenAI's product record. ChatGPT ads went live in the UK on June 6, 2026, and the self-serve Ads Manager opened to UK advertisers on June 22. Conversion-optimised campaigns began rolling out on June 5, 2026, and by August OpenAI had made a Maximize results automated bidding strategy the default for eligible new ad groups, able to optimise toward conversions. On October 5, 2026, two days before the Brainlabs post, OpenAI listed more than 20 measurement partners for ChatGPT Ads, covering attribution, geo experiments and brand studies. Availability may vary by market and account, which the post does not address.

Measurement gaps

Brainlabs calls measurement "the part nobody has solved." AI referral visits, it says, are regularly recorded as direct traffic, and "everything quoted in this piece, our own figures included, probably understates the real scale."

One sentence in this section is hard to assess. The post states that "Google states that ChatGPT product considerations 99% of the time result in Google Searches for purchase decision making." No source, date or document is given for the claim, and the wording does not make clear what was measured. Without a citation, it cannot be verified, and it is not treated as a Google statement here.

The post names two responses. One is giving AI referrals their own channel in GA4 so their conversion rate is visible separately. Google has partly done this already: on May 13, 2026, Google Analytics added an "AI Assistant" channel to the Default Channel Group, classifying sessions from ChatGPT, Gemini and Claude automatically when the referrer is recognised. That does not catch the visits that arrive with no referrer, which is the problem Brainlabs describes.

The other response is a holdout group, which the post calls "the only way to show what ChatGPT ads caused." That is a statement about incrementality: attribution tools can show which clicks preceded a sale, but only a withheld control group shows what would have happened without the ads. According to Brainlabs, both measures are "far easier to build before finance asks for a clean number than after."

Why this matters for search marketers

The Brainlabs post adds agency account data to a debate PPC Land has followed for more than a year. Its organic decline of 10.5% is milder than several published estimates. Ahrefs found in February 2026 that AI Overviews were associated with a 58% drop in click-through rate for top-ranking pages, up from 34.5% in April 2025. On the paid side, Seer Interactive reported in November 2025 that paid CTR on informational queries with AI Overviews fell 68%, from 19.7% to 6.34%, between June 2024 and September 2025.

The claim that AI-referred visitors convert better also has company. Microsoft Clarity found in November 2025 that AI referrals converted to sign-ups at 1.66%, against 0.15% for search, across 1,277 publisher domains. Brainlabs' own July 2025 report put the value of AI search visitors at 4.4 times that of traditional organic traffic. Google has made a version of the same argument for its own product, with John Mueller and Elizabeth Reid describing AI Overview clicks as "higher quality" in 2025.

Not all the evidence points one way. A study of 973 e-commerce sites with about $20 billion in combined revenue, covering August 2024 to July 2025, found that ChatGPT traffic converted below every channel except paid social, and that it made up less than 0.2% of sessions. Brainlabs' own monthly table shows AI referrals under-converting organic search until early 2026. The direction may have changed, but how recently it did is relevant to anyone using these figures to justify budget.

Scale is the other constraint. The Brainlabs data puts AI referrals at about 1% of combined sessions. A Smarty Marketing survey of 1,295 US consumers in May 2026 found that 56.68% start product searches on Google and 7.26% on ChatGPT. The post cites ChatGPT passing 1 billion monthly active users in May 2026, and says roughly 70% of people using it for shopping-related tasks turn to it for research and discovery. It does not give a source for either figure.

The keyword problem the post describes is not limited to ChatGPT. In May 2026, Google stated that for searches from Lens, AI Mode, AI Overviews and auto-complete, the search term in reporting is the "best approximation of the user's intent" rather than the literal query. The Brainlabs post calls keywords "always a workaround" and describes match types, negative keyword lists and search term reports as "tools for guessing at intent from a handful of words." Google's own reporting changes suggest the platform is moving away from the literal keyword too.

The post closes on timing. According to Brainlabs, "the brands that get there first, while the format is still rough and the CPMs haven't caught up to Google's, are the ones who'll set the terms for everyone who tests it after them." That is a forecast from an agency that sells this work. The data in the post supports a narrower conclusion: across its clients, organic and Search click volume are shrinking, AI referrals are growing from a small base, and those visits have recently started converting at a higher rate.

Timeline

Summary

Who: Brainlabs, a paid media and SEO agency, in a post by Alessio Messaggi, drawing on data from 54 clients for organic traffic and 31 Google Ads accounts for spend and click mix.

What: The agency argues that buyers increasingly compare options inside AI assistants before clicking an ad, so clicks arrive later in the decision and convert at higher rates. Its data shows organic sessions down 10.5%, AI referrals replacing about 5% of the loss, Search campaigns' share of clicks falling from 62% to 41%, and AI referral key event rates of 13.4% against 10.4% for organic. It also reports ChatGPT ad CPMs of £15 to £60. Some figures, including a 1.5 times conversion gap and a claim attributed to Google about 99% of ChatGPT product considerations, cannot be verified from the material published.

When: The post was published on October 7, 2026. The click mix data runs from January 2024 to August 2026, and the monthly organic and AI referral data from January 2025 to April 2026.

Where: The post appeared on the Brainlabs blog. The client data spans the US, UK and other markets, and the ChatGPT pricing is quoted in pounds without a stated market.

Why: The analysis matters for search marketers because click volume, the metric paid search has long been planned around, is falling in agency data while the value per click from AI-referred visitors is rising, and the conversations where buyers make their choices sit outside the reach of keyword targeting and most measurement tools.