A growing thread on LinkedIn captures a frustration that has been building among paid-search practitioners for almost a year: AI Max for Search campaigns, the format that will absorb Dynamic Search Ads in September 2026, is failing at the one job DSA reliably performed - sending granular product queries to the matching product page.

Joey Bidner, a freelance Google Ads manager, posted a side-by-side example on LinkedIn in early May 2026 showing three specific search terms that, in a Dynamic Search Ads campaign, would have routed shoppers directly to individual product pages. Two of the queries included exact refrigerator sizes. One included a product identifier. In AI Max, according to the screenshots Bidner shared, all three went to the same generic page.

"Here is an example of why AI max SUCKS compared to DSA," Bidner wrote in the post. "These are 3 very specific search terms that, in a DSA campaign, would send users directly to the specific product page. Two include exact fridge sizes, and one even has a product ID. That's as granular as it gets. DSA typically nails this and routes users straight to the exact product since those IDs and size details are live on the product pages.... and with AI Max..... THEY ALL JUST GO TO ONE BASIC PAGE. HOW IS THIS BETTER THAN DSA Ginny Marvin!"

The post drew 58 reactions, 23 comments, and 2 reposts within roughly two weeks, including a direct response from Ginny Marvin, Google's Ads Product Liaison. Her reply acknowledged the technical scope of the complaint without disputing the underlying observation. According to Marvin, "AI Max does support using a combination of URL rule types (URLs, custom labels in page feeds, URL combo rules), similar to DSA. URL exclusions are managed at the campaign level and URL inclusions are managed at the ad group level. AI Max also supports page feeds. With page feeds, you can add multiple custom labels."

Marvin also confirmed that the new system does not yet support every rule type DSA accepted. "AI Max doesn't support all of the rules you may be currently using (e.g. page contains) in DSA," she wrote in the LinkedIn thread. "For existing DSA ad groups, the new migration tools will transfer all existing URL settings. Any rule combinations or targets that aren't available in AI Max will be brought over as read-only and continue to work. They can't be edited though."

That admission - that legacy targeting rules will continue to function but cannot be modified after the transition - sits at the centre of the practitioner concerns now circulating publicly.

The technical mismatch advertisers describe

Dynamic Search Ads have operated on a defined mechanic since their introduction. Google AdsBot crawls an advertiser's website, indexes landing-page content, and matches incoming search queries to the page whose content most closely aligns with the query. Targeting was configurable through page feeds, custom labels, URL rules, and content-based filters such as "page title contains" or "page content contains."

AI Max for Search campaigns - introduced on May 6, 2025 and declared out of beta on April 15, 2026 - operates on a broader input set. According to the Google Ads help documentation for AI Max setup, the system draws on landing pages, existing ad copy, keywords within the ad group, and real-time intent signals. Three settings drive its behaviour: search term matching, text customization, and Final URL Expansion.

Final URL Expansion is the component most directly relevant to the LinkedIn complaint. According to Google's documentation on the feature, Final URL Expansion is opt-in and enabled by default when AI Max is turned on. The documentation describes it as a campaign-level setting designed to "send traffic to the most relevant URLs on your domain when it's likely to result in better performance." It further specifies that "Final URL expansion will only send traffic to URLs that are query relevant and themed to your ad group."

The system relies on Text customization being enabled. According to the documentation, "In order to use the Final URL expansion setting, Text customization must be selected. Turning off Text customization will also disable Final URL expansion." That dependency means an advertiser who wants to constrain creative generation cannot, by design, retain the URL routing layer that AI Max provides.

Bidner's follow-up explained why the published controls did not solve his issue. "The issue here is less about exclusions and rules, its about page matching," he wrote in the thread. "Even without any specific rules in DSA, just simple targets, the system can very accurately match a granular query to a specific product page it matches, where ai max simply doesnt. 90% of searches go to a gerneric page. these 3 searches in the screen shot are perfect examples. in my dsa, those searches go to their specific pages and that is not with any complex set ups, just a simple theme target."

Marvin's response to that follow-up was brief. "Yes, I understand what you are saying and the example you shared, and have passed the feedback to the team."

Exclusion behaviour in question

Rameez Ramzan, a performance marketing manager commenting on the same thread, raised a separate but related concern about URL exclusions in AI Max. "URL exclusions - In DSA it works spot on, but in AI Max it is not accurate," he wrote. "We excluded URLs under the asset optimization setting, but the campaign still targets those pages."

The URL exclusion mechanism in AI Max is governed by the same documentation page that defines Final URL Expansion. According to that page, URL exclusions allow advertisers to block specific pages from serving through Final URL Expansion, but the exclusions are dependent on Text customization and Final URL Expansion both being turned on. URL inclusions, by contrast, are an ad-group-level control intended to direct delivery to a list of specified pages.

Scott Clark, a digital marketing consultant, noted in the thread that weak product detail pages or cannibalization issues might cause similar behaviour in DSA. Bidner pushed back. "For sure, but in this case, I'm talking about an account with good pages," he replied. "We get these same searches go to specific product pages with dsa no problem. I'm running ai max to try and get it working in time for their stupid 'upgrade' cut off date and it's just been a massive disappointment."

The September 2026 deadline

The frustration intensified by the timeline Google has committed to publicly. Google announced on April 15, 2026 that Dynamic Search Ads will be retired as a standalone format, with all eligible legacy campaigns automatically upgrading to AI Max for Search starting in September 2026. The announcement was published by Brandon Ervin, Director of Product Management at Google Ads, on the Google Ads and Commerce Blog.

Three legacy configurations are affected: Dynamic Search Ads, automatically created assets, and the campaign-level broad match setting. New DSA campaigns cannot be created via Google Ads, Google Ads Editor, or the Google Ads API from September onward. According to the Google Ads page feed documentation, "To help more advertisers access the benefits of AI Max, starting in September 2026, campaigns using Dynamic Search Ads (DSA), automatically created assets (ACA) and campaign-level broad match setting will automatically be upgraded to AI Max."

For DSA advertisers specifically, the migration converts dynamic ad groups into standard ad groups. All three AI Max features - search term matching, text customization, and Final URL Expansion - are switched on by default, with existing URL controls preserved. After the upgrade, parallel running maintains legacy Dynamic Search Ads temporarily while the system generates new Responsive Search Ads and submits them to policy review. When the new RSAs pass policy checks and become active, the legacy DSAs are paused and placed in a "view-only" state.

According to Google's documentation, for new page feed uploads or edits, "it can take an average of 2-14 days for Google to crawl the feed (depending on the feed size)." That window has practical implications during the transition period for advertisers who manage feeds with thousands of URLs.

A pattern of practitioner concern

The May 2026 LinkedIn exchange is not the first time AI Max performance has been publicly contested. Independent testing has repeatedly found gaps between Google's headline claims and observed outcomes.

In August 2025, Ezra Sackett, Director of Paid Search at Monks, reported that "99% of impressions have 0 conversions" across approximately 30,000 search terms in client accounts that activated AI Max features. In November 2025, Smarter Ecommerce published an analysis of more than 250 retail campaigns showing AI Max delivering conversions at approximately 35 percent lower return on ad spend than traditional match types. A four-month test in the same period showed a cost per conversion of $100.37 under AI Max compared with $43.97 for phrase match.

Earlier in August 2025, industry analysts flagged aggressive expansion of AI Max onto Search Partner Network placements. Mike Ryan, Head of Ecommerce Insights at Smarter Ecommerce, described AI Max behaviour on SPN placements as "expansionary" and "deeply disturbing." Research from Intelligency Group cited in the same coverage found that SPN placements delivered 37 percent lower return on ad spend than Google Search proper.

The performance figures Google itself cites have shifted. At launch in May 2025, the company promised 14 percent more conversions or conversion value at similar CPA or ROAS. At the time of the April 2026 DSA retirement announcement, that figure had been revised to 7 percent more conversions when using the full AI Max feature suite compared with search term matching alone.

Reporting and visibility

Search term reporting has been another sustained source of friction. Google clarified in December 2025, following questions raised by advertising professional Brad Geddes, that AI Max determines relevance through inferred user intent rather than raw text matching. The clarification addressed search term reporting anomalies that had surfaced in advertiser accounts.

That distinction - between literal query matching and inferred-intent matching - is structurally relevant to Bidner's complaint. When an advertiser sees "dec3050r" or a specific fridge dimension as a search term, the expectation built into a decade of DSA practice is that the system will route the click to the page whose content most directly matches that string. AI Max, by design, may route the same query to a page the system has inferred is more likely to convert based on aggregate intent signals rather than the literal contents of the query.

Practitioners have also pointed to the trajectory of Google's automation strategy. Google ended Dynamic Search Ads at a moment when AI-mediated matching is becoming the platform's default targeting infrastructure. The text guidelines feature, which went global on February 26, 2026, gives advertisers up to 25 term exclusions and 40 messaging restrictions to constrain AI-generated ad copy. That feature governs what ads say, not what queries trigger them - a distinction that underscores the separation between creative controls, which have expanded, and targeting transparency, which has narrowed.

The control layer

Several commenters in the LinkedIn thread questioned the rationale for retiring DSA at all. Chris Chambers, Head of Paid Search at Understory, wrote: "Ya it doesn't make much sense to me, to actually kill a keywordless campaign for a campaign they say is keywordless yet it still has keywords. Google doesn't make sense sometimes."

Paul DeMott, a digital growth strategist, framed the timing concern directly: "It's a shame DSA is going so soon before AI Max is running cleanly." Bidner replied that he hoped the situation would improve before the deadline. "Hopefully it will get MUCH better before the cut off date for DSA. pmax got better, so there is hope..."

The frustration extends to upcoming controls that Marvin mentioned but that are not yet live. According to her reply in the thread, Google is "aiming to bring the content and titles related exclusions to the account level later this year. This will complement AI Max's current inventory aware serving that automatically excludes out-of-stock items. It will give you the ability to always exclude any other content you don't want to use in your ads at the account level."

Inventory-aware serving and account-level exclusions address related but distinct problems from the page-matching issue Bidner raised. Excluding out-of-stock items prevents wasted clicks on unavailable products. The unresolved concern is not preventing inappropriate matches but ensuring that valid, granular queries reach the corresponding product pages.

What changes operationally

For DSA advertisers preparing for the September 2026 cutoff, the Google Ads help center documentation outlines a specific upgrade path. The procedure runs through the Campaigns menu, with a callout to upgrade appearing under the Dynamic Search Ads settings panel for affected campaigns. After confirmation, dynamic ad groups convert to standard ad groups, Dynamic Search Ads become Responsive Search Ads with the minimum static assets required, and Final URL expansion can be enabled.

Legacy targeting rules such as "page title contains" or "page content contains" are flagged in the confirmation dialog as becoming read-only after the upgrade. New rules of those types cannot be added. Existing rules can be removed but not edited. The migration tools, according to Marvin's LinkedIn explanation, preserve all existing URL settings. Unsupported rule combinations are brought over as read-only and continue to function.

That preservation creates an awkward intermediate state. Accounts can keep their existing logic but lose the ability to modify it. Any new optimization must work within the AI Max framework rather than the DSA framework. Over time, as websites evolve, the read-only DSA rules will reference pages, categories, and content that no longer match the site as it exists, with no available repair path.

Industry significance

The exchange between Bidner and Marvin is unusual in that it places a granular, technically specific complaint about Google Ads in front of a senior Google representative in a public forum, and elicits a direct acknowledgement that the system does not yet replicate the function the practitioner relies on. That kind of public exchange is rare in a category dominated by official documentation and platform-managed communication channels.

PPC Land has documented advertiser concerns about AI Max throughout the product's first year, from the August 2025 Search Partner Network expansion findings through the November 2025 ROAS analysis and the December 2025 attribution clarifications. The pattern has been consistent: independent testing surfaces a specific deficit, Google introduces a control feature or reporting enhancement intended to address it, and the underlying targeting and routing behaviour remains a source of disagreement.

The September 2026 cutoff is the point at which that disagreement becomes operationally binding. Advertisers who have relied on DSA's granular routing logic - particularly retailers with deep product catalogs, lead generation accounts with topic-specific landing pages, and multi-brand portfolios with distinct site sections - will have their campaigns migrated to a system that, in the assessment of practitioners actively testing it, does not yet replicate the routing behaviour they have built workflows around.

Timeline

Summary

Who: Joey Bidner, a freelance Google Ads manager, posted the original LinkedIn thread that surfaced the routing complaints. Other commenters included Rameez Ramzan, Scott Clark, Paul DeMott, Chris Chambers, Ibrahim Kurmywal, Shubham Soni, Joseph Williams, Jordana Keely, and Tyler Rabey. Ginny Marvin, Google's Ads Product Liaison, responded directly in the thread. The system at issue, AI Max for Search campaigns, is a product of Google Ads under Brandon Ervin, Director of Product Management at Google Ads.

What: Practitioners are publicly contesting the page-matching behaviour of AI Max for Search campaigns, arguing that the system routes granular, specific product queries - including queries containing product IDs and exact specifications - to generic landing pages rather than the matching product pages. The same accounts running Dynamic Search Ads continue to route those queries accurately. The complaint focuses on the Final URL Expansion mechanism, which is enabled by default when AI Max is activated and which depends on Text customization being enabled to function.

When: The LinkedIn thread surfaced approximately two weeks before publication, with the underlying technical concerns building since AI Max launched in May 2025. The operational deadline driving the urgency is September 2026, when Dynamic Search Ads will be retired and remaining eligible campaigns automatically upgraded to AI Max.

Where: The exchange took place on LinkedIn. The underlying campaigns affected are Google Ads accounts globally, with the AI Max framework operating across Google Search, Google Ads Editor, and the Google Ads API.

Why: The page-matching gap between DSA and AI Max is particularly relevant for advertisers managing accounts with granular product taxonomies, where exact-match routing materially affects conversion outcomes. Google's documentation positions AI Max as a system that draws on a broader input set than DSA - landing pages, ad copy, keywords, and real-time intent signals combined - but practitioners testing the system in live accounts report that this broader input has not translated into better page selection for specific queries. With Dynamic Search Ads being phased out in September 2026 and no comparable replacement available outside the AI Max framework, advertisers face the prospect of losing a routing precision they have built workflows around for years.

Share this article
The link has been copied!