Retail search campaigns running Google AI Max were exposed to 72% more invalid traffic than search campaigns without it, according to a report published today by invalid traffic detection company Lunio. The finding lands nineteen days before Google converts two legacy campaign settings to AI Max automatically.

The study covers more than 414 million clicks recorded across Google Ads, Bing, LinkedIn, Meta and leading native and social platforms between October 2025 and June 2026, according to Lunio. It is the latest instalment in a series of vertical-specific invalid traffic studies the company has published through 2026, this one narrowed to retail brands.

The headline comparison is internal to Google search. Invalid traffic rates on retail search campaigns with AI Max enabled more than doubled over the observation window, climbing from 2.46% in the fourth quarter of 2025 to 5.28% in the second quarter of 2026, according to the report. Standard search campaigns moved in the opposite direction over the same three quarters, easing from 3.72% to 3.07%.

Two lines diverging is a different kind of finding from a single rising number. One campaign configuration deteriorated. The other, running on the same platform, in the same accounts, against the same advertiser base, did not.

What the dataset covers

Lunio defines invalid traffic, abbreviated to IVT, as any click, impression or conversion that does not originate with a person holding genuine intent to engage. The category spans coordinated bot activity, automated scraping, clicks manufactured by competing advertisers, and accidental clicks.

Across the full retail sample, brands recorded an average IVT rate of 5%. That figure conceals a quarterly trend. Rates sat at 4.18% in the fourth quarter of 2025 and reached 5.54% in the second quarter of 2026, the highest reading anywhere in the dataset, according to Lunio.

The measurement window matters for interpretation. October 2025 through June 2026 spans one complete holiday trading season and the quieter months that follow it. The peak reading arrives in the second quarter of 2026, outside the seasonal spend concentration rather than inside it.

The AI Max share of invalid clicks

Beyond the rate comparison sits a distribution figure. Across the entire Google search campaign dataset, AI Max accounted for more than two-thirds of all invalid clicks detected, or 68%, according to the report. Lunio adds that these invalid clicks more than doubled during the observation period.

Share of invalid clicks and rate of invalid clicks are separate measurements, and the gap between them is where interpretation gets difficult. A campaign type can carry a large share of a problem simply by carrying a large share of the traffic. What makes the 68% figure harder to dismiss on volume grounds alone is the parallel rate movement: if AI Max were merely absorbing traffic that would otherwise have flowed through standard search at comparable quality, the two rate lines would track each other. They do not.

Google averaged a 4.75% IVT rate across all campaign types combined, according to Lunio. Every channel measured within the Google set recorded a higher rate at the end of the period than at the start.

Shopping records the sharpest climb

Among Google campaign types, Google Shopping carried the highest average IVT rate across the sample period at 6.33%, followed by Display at 6.05% and Search at 4.73%.

Shopping also produced the steepest trajectory. The rate moved from 4.16% in the fourth quarter of 2025 to 7.51% in the second quarter of 2026, an increase of 80%. By the end of the period, roughly one in thirteen Shopping clicks was invalid, according to the report.

That campaign type is the one most distinctively retail. Merchants without a Shopping presence are the exception rather than the norm in the vertical, and Shopping inventory sits closest to the transaction. A click on a product listing with a visible price carries a narrower band of possible intent than a click on a text advertisement, which is part of why the channel has historically been treated as high quality and priced accordingly.

Google extended the AI Max framework to standard Shopping campaigns on April 30, 2026, introducing automated features that match product advertisements against conversational queries rather than exact product lookups. That change lands inside the measurement window, though the report does not attribute the Shopping movement to it.

Where retail money meets platform rates

Outside Google, Meta recorded the highest rate of any major retail channel at 5.99% on average, peaking at 6.81% in the first quarter of 2026. TikTok averaged the second highest rate at 5.56%, with invalid traffic up 68% across the measurement period, according to Lunio.

The ordering is worth setting against Lunio's earlier work. The company's Invalid Traffic Impact Report covering the IT and security sector, published in June 2026 and drawn from 64 million clicks, found LinkedIn recording 17.62% in the first quarter of 2026, the highest platform-level figure in that study. A subsequent Lunio report on banking, lending and credit found Google Search carrying 5.51% for that category, inverting the usual pattern in which Search is the cleanest Google channel.

Different verticals, different channel mixes, different results. The consistent element across all three studies is directional: rates rose quarter on quarter in each.

The cost model

Lunio translates the retail rate into money using a worked example rather than an aggregate market estimate. For a retail business spending $10 million per year on paid advertising at an average cost per click of $3.70, the 5% rate corresponds to roughly $500,000 in directly wasted advertising spend annually, according to the report.

Applying a conservative three-to-one return on advertising spend ratio, Lunio puts the lost revenue opportunity at approximately $1,250,000 each year for that same business.

The arithmetic is transparent and its limits are visible. The figure scales linearly with spend, which means a retailer running a $2 million budget arrives at a fifth of it. It also assumes the blocked traffic would have converted at the account average had it been real, which is the standard convention in wasted-spend modelling and the standard point of contention with it.

Nineteen days to the auto-upgrade

The report arrives at a specific point in the AI Max deployment schedule. Google Ads emailed advertisers on August 5, 2026 confirming that campaigns running automatically created assets or the campaign-level broad match setting will be converted to AI Max for Search campaigns starting September 1. The notice carried no accompanying blog post and was signed by the Google Ads Team rather than a named executive.

For the broad match cohort, that conversion is close to a relabelling. For the automatically created assets cohort, it is not. Automatically created assets governed creative generation. Search term matching governs which queries an advertisement becomes eligible for. Campaigns that opted into automated headline writing will inherit automated query expansion unless the setting is switched off at ad group level.

The wider migration began earlier. Google announced on April 15, 2026 that Dynamic Search Ads would be retired as a standalone format and declared AI Max out of beta. That deadline later moved: a rewrite of the AI Max reporting documentation in June 2026 pushed the DSA timeline to 2027. The September 1 date for automatically created assets and campaign-level broad match survived unchanged.

Nick Morley, CEO of Lunio, tied the retail exposure to the shape of the trading calendar rather than to any single platform decision.

"With many retailers relying on automated media buying and the prevalence of sales events that compress enormous spend into short budgets, retailers are particularly vulnerable to the risks and costs associated with invalid traffic," Morley said, according to the report. "As Google AI Max continues to play a larger role in media buying for retailers, especially ahead of the all-important holiday shopping season, retailers need to audit their traffic quality and ensure that they're feeding higher quality traffic into their campaigns to optimize against higher-intent visitors during this pivotal season."

A performance record already under scrutiny

AI Max entered open beta on May 6, 2025 carrying a claim of 14% more conversions or conversion value at similar cost. Independent measurement has run against that number from the outset. In August 2025, analysis from Ezra Sackett, Director of Paid Search at Monks, found that 99% of AI Max impressions produced zero conversions across roughly 30,000 search terms.

A larger study followed. Smarter Ecommerce examined more than 250 retail campaigns and published findings on November 6, 2025 showing AI Max delivering conversions at roughly 35% lower return on advertising spend than traditional match types inside the same campaigns, with higher cost per conversion and lower average order values. The same firm reported this month that exact match keywords are losing about one in four impressions to AI Max.

Google revised its own headline figure in the April 2026 general availability announcement, citing an average of 7% more conversions or conversion value across the full feature suite compared with search term matching alone, based on internal 2026 data that excluded retail advertisers.

The Lunio findings add a distinct variable to that record. Return on advertising spend gaps can be explained by matching quality, by attribution shifts, by incrementality questions or by the composition of expanded traffic. An invalid traffic rate is a claim about whether the clicks came from people at all.

The measurement layer underneath

Attribution complicates any attempt to reconcile the two. PPC Land documented in December 2025 that AI Max reporting credits conversions on inferred intent rather than literal query text, which makes it difficult for advertisers to separate traffic AI Max generated from traffic it claimed. If the credit assignment is opaque, the quality assessment inherits that opacity.

Query-level visibility has improved in stages. Branded search controls arrived inside AI Max in 2026, offering a native toggle governing whether advertisements appear on queries containing brand names. Search term match type segmentation was added to the Keywords tab in 2025, alongside flagged concerns about Search Partner Network expansion. None of those controls report on traffic validity.

That gap is not unique to Google. Lunio's own survey work, published on July 15, 2026, found that 5.3% of 131 senior marketers run a dedicated invalid traffic platform while 75.6% estimate losing more than 5% of monthly performance budget to bots. The company labels the distance between the two the 95% execution gap.

Verification vendors have reported movement in the other direction. DoubleVerify said on July 29, 2026 that fraud and invalid traffic violation rates fell 41% year over year in North America and 45% in Europe, the Middle East and Africaamong protected campaigns, while in unprotected media AI bots generated up to ten times more clicks than humans in some studies. Both sets of figures can hold simultaneously, since they measure different populations: campaigns running detection, and campaigns running without it. Lunio's retail data comes from client accounts, and the company has previously described its measurement samples as operating in monitor-only mode, tracking invalid traffic without blocking it.

Why this matters for retail advertisers

The fourth quarter compresses a disproportionate share of annual retail advertising spend into a short window, which raises the absolute cost of any percentage rate. Integral Ad Science documented in July 2026 that traffic to made-for-advertising and ad clutter sites rose 5% across Christmas Eve and Christmas Day 2025, arguing that low-quality inventory volume increases precisely when advertisers spend most. The Lunio retail dataset, which peaks after the holiday period rather than during it, does not test that seasonal thesis directly, and the two studies measure different phenomena.

What the two do share is a timing problem. Rates measured in a quieter quarter get applied to budgets deployed in a louder one. A 5.28% AI Max rate against a $10 million annual budget is one number. The same rate against fourth-quarter spend concentrated into six weeks is the same percentage attached to a much larger denominator per unit of time.

The September 1 conversion adds a second variable. Advertisers whose campaigns move to AI Max automatically will enter the holiday period on a configuration that Lunio measured at nearly double the invalid traffic rate of the setting it replaces, on a dataset that ends in June. Whether that relationship persists at seasonal volume is not something any published measurement currently answers.

Timeline

Summary

Who: Lunio, an invalid traffic detection and prevention platform, with findings attributed to Nick Morley, chief executive of the company. The report addresses retail advertisers running paid media across Google, Bing, LinkedIn, Meta and leading native and social platforms, and bears on any retailer whose campaigns fall under the September 1, 2026 AI Max conversion.

What: A retail-specific invalid traffic report finding that Google search campaigns with AI Max enabled were exposed to 72% more invalid traffic than search campaigns without it, that AI Max accounted for 68% of all invalid clicks detected across the Google search dataset, and that retail invalid traffic averaged 5% overall. Google Shopping carried the highest campaign-type average at 6.33% and the sharpest climb, reaching 7.51% in the second quarter of 2026. Meta recorded 5.99% and TikTok 5.56%. For a retailer spending $10 million annually at a $3.70 average cost per click, Lunio models roughly $500,000 in wasted spend and approximately $1,250,000 in lost revenue opportunity at a three-to-one return ratio.

When: The report was published today, August 12, 2026. Its measurement window runs from October 2025 to June 2026, covering the fourth quarter of 2025 through the second quarter of 2026.

Where: Released from New York. The analysis covers retail brand accounts across Google, Bing, LinkedIn, Meta and leading native and social platforms, with campaign-type breakdowns for Google Shopping, Display and Search.

Why: The findings arrive nineteen days before Google automatically converts campaigns using automatically created assets or the campaign-level broad match setting to AI Max for Search, and roughly three months before the fourth-quarter retail trading peak, when a percentage rate applied to compressed budgets translates into a larger absolute cost per unit of time.