Lunio Report Links AI Max Adoption to Higher Invalid Traffic in Retail Search

Posted on Oct 26 • Originally published at blog.geosearchleads.com

Lunio has published a retail-focused invalid traffic study that raises a significant measurement concern for advertisers using Google AI Max. Its analysis of more than 414 million retail advertising clicks found that invalid traffic rose across the sector between late 2025 and mid-2026. In Lunio’s observed Google Search campaigns, AI Max was associated with materially higher invalid traffic rates than standard Search.

The findings do not establish that every AI Max campaign will experience the same result, or that all questionable clicks are fraudulent. They do, however, show why retailers should assess traffic quality before treating additional clicks, reach, or automated campaign expansion as evidence of stronger performance.

According to Lunio’s Retail Invalid Traffic Impact Report, average invalid traffic (IVT) across the retail campaigns in its dataset was 5.00%. The analysis covered October 2025 through June 2026 and included activity across Google, Meta, TikTok, and Bing.

Lunio reported that the average rose over the period, from 4.18% in Q4 2025 to 5.54% in Q2 2026. Individual campaigns had substantially larger exposure, with IVT as high as 81% in the dataset. That range is important: an average rate can help benchmark a portfolio, but it cannot identify whether a particular account, campaign, query group, or inventory source is affected.

The report uses a straightforward budget illustration. At a 5.00% IVT rate, a retailer spending $5 million annually on advertising could lose roughly $250,000 in direct media spend. Lunio further models approximately $750,000 in lost opportunity using a conservative 3:1 return on ad spend. This is an example, not a universal estimate of retail losses, but it illustrates how a small percentage of low-quality paid clicks can become commercially meaningful at scale.

Channel-level results also varied. Lunio reported approximately 5.99% IVT for Meta and 7.51% for Google Shopping during the study period. Overall Google-channel IVT sat in the mid- to upper-4% range depending on the subchannel.

The most consequential part of the research concerns Google AI Max, the campaign setting designed to use AI-driven targeting, matching, and creative capabilities in Search campaigns. Lunio compared campaigns from the same customer base across a late-2025 to mid-2026 window.

In that like-for-like analysis, Lunio said IVT rose from 2.46% when AI Max was off to 5.28% when it was on. That is a 114% increase based on those two rates. Its supporting analysis also reported that IVT in standard Search declined by about 17% during the same period.

By the end of Q3 2026, AI Max accounted for around 55% of retail Search clicks in Lunio’s data, while representing 68% of invalid Search clicks. The research therefore suggests that increasing AI Max adoption coincided with a disproportionate share of questionable traffic in the observed accounts. Lunio’s AI Max invalid-traffic analysis provides further detail on the comparison and the reported share of invalid clicks.

Google began automatically upgrading eligible legacy Search campaigns to AI Max from September 2026, according to the research. That makes traffic-quality controls more urgent for retailers that have not deliberately evaluated the setting, particularly where legacy campaign data is used as a benchmark for budget decisions.

Automation can broaden the combinations of queries, assets, and inventory involved in serving ads. That may generate incremental exposure that a retailer would not have reached through a narrower setup. Lunio explicitly notes that some IVT may be incremental because of expanded inventory, rather than fraud alone. The operational issue remains the same: advertisers need to determine whether additional clicks produce valid sessions, engaged users, and meaningful business outcomes.

This also affects search measurement beyond paid media. Weak paid traffic can distort audience signals, landing-page engagement metrics, assisted-conversion models, and decisions about which product categories or queries deserve more content investment. Organic search and AI-discovery teams should not interpret paid click volume as direct evidence of genuine demand without considering traffic quality and downstream behavior.

The report is observational, not proof of causation. Advertisers opted into AI Max, and the results may not apply to every retailer, campaign type, or industry. Still, the findings support a more disciplined approach to testing.

Start by separating AI Max-enabled campaigns from standard Search in reporting. Compare more than click-through rate and cost per click. Review the full sequence from ad click through to server-side or analytics session, product view, add-to-cart event, checkout, and validated conversion where available.

The goal is not to assume all automated expansion is harmful. It is to establish whether incremental traffic is commercially valid before allowing automated systems to scale spend against it.

For teams connecting paid-search data with wider AI-search and organic visibility work, a site-level baseline can help separate genuine discoverability opportunities from noisy acquisition signals. Create a GEO Search Leads account to assess your website’s GEO readiness and identify optimization opportunities that support clearer, more reliable visibility planning.

What is invalid traffic in retail advertising?

Invalid traffic is ad traffic that does not represent a legitimate potential customer interaction. Lunio’s study measures IVT across retail ad campaigns, but the supplied research does not provide a complete breakdown of every underlying IVT type.

Did Lunio prove that AI Max causes invalid traffic?

No. Lunio’s analysis is observational and compares advertisers that opted into AI Max with campaigns from the same customer base and period. It identifies an association in the observed data, not universal causation.

How much higher was invalid traffic with AI Max in Lunio’s comparison?

Lunio reported 5.28% IVT with AI Max enabled compared with 2.46% with AI Max off in its like-for-like Google Search comparison. That is a 114% increase based on those rates.

Does a high click volume mean a retail campaign is performing well?

Not necessarily. Click volume should be assessed alongside validated sessions, engagement, conversion quality, and business outcomes. The study shows why click counts alone can be an incomplete performance measure.

Lunio’s retail analysis does not make AI Max inherently unsuitable for every advertiser, but it provides a clear reason to test it with stronger traffic-quality controls. As AI-driven campaign expansion becomes more common, retailers should evaluate incremental clicks by their validated value, not just by the volume and efficiency metrics shown inside an ad platform.