Google Shopping’s AI Max Campaigns Upend Merchant Bidding in 2026

When Google quietly rolled out AI Max for Performance Max campaigns in late March 2026, the search giant framed it as a routine quality-of-life upgrade — better asset matching, smarter audience expansion, tighter search-term controls. But for the Shopify operators and DTC brands managing six- and seven-figure Google Shopping budgets, the shift has been anything but routine. Five months in, merchants are reporting wildly split outcomes: some are posting their lowest cost-per-acquisition numbers in three years, while others have watched ROAS collapse almost overnight.

The divergence is creating a sharp skills gap between brands that know how to feed Google’s machine and those still operating on pre-2025 campaign instincts.

What exactly is AI Max, and how does it change Google Shopping campaigns?

AI Max is effectively a new signal layer inside Performance Max that gives Google broader latitude to match search queries to product listings, expand audience targeting beyond first-party seed lists, and dynamically rewrite ad copy at the asset group level. For Shopping-heavy advertisers, the most consequential change is the expanded search-term matching: AI Max can serve a product listing against queries that have no direct keyword overlap with the product feed, relying instead on inferred intent signals.

Thomas Pringle, head of paid search at Grove Commerce, a Shopify-focused performance agency managing roughly $40 million in annual Google ad spend, says the mechanics require a fundamentally different setup discipline.

“The old PMax playbook was about tightly segmented asset groups and a clean feed. AI Max blows the lid off query matching, which means if your product titles and descriptions are weak, you’re now surfacing against totally irrelevant intent. Feed quality isn’t just a nice-to-have anymore — it’s the single biggest lever you have.”

💡 Article Summary

Key Insights

1

What exactly is AI Max, and how does it change Google Shopping campaigns?

2

Which merchant categories are seeing the biggest gains — and who is getting hurt?

3

How are agencies and in-house teams adapting their campaign structures?

4

What does AI Max mean for attribution and blended CAC reporting?

5

Should merchants pause AI Max or lean further in?

Source: Ecommerce Times

Pringle’s agency has responded by running Feedonomics-optimized feeds with custom labels segmenting by margin tier, seasonal velocity, and new-customer eligibility — a setup that costs an additional $1,200 to $2,500 per month in feed management overhead but has, in his telling, driven a 28% improvement in new-customer ROAS across three apparel clients since April.

Which merchant categories are seeing the biggest gains — and who is getting hurt?

The performance split tracks closely with average order value and catalog complexity. High-AOV merchants in home goods, outdoor equipment, and specialty apparel are disproportionately benefiting from AI Max’s expanded query matching, because a single misfire on a $15 item costs very little, but surfacing against a high-intent adjacent query for a $400 item can be transformative.

Mara Delacroix, founder of Portland-based outdoor gear DTC brand Ridgeline Supply Co., says her Google Shopping ROAS climbed from 4.1x to 5.6x in the six weeks after her agency enabled AI Max with a fully segmented feed.

“We were spending $85,000 a month on Google and leaving a lot of category traffic on the table because we were too conservative with our match types. AI Max found buyers we were never going to reach with manual Shopping campaigns. Our average ticket is $310, so every incremental conversion matters.”

The picture is grimmer for low-margin, high-volume merchants — particularly those selling in commoditized categories like consumer electronics accessories, phone cases, and generic home essentials. Several sellers in Ecommerce Times’ reporting described AI Max expanding their reach into informational and comparison queries where conversion rates are structurally low, burning budget without generating meaningful return.

• Consumer electronics accessory sellers on Shopify reporting average ROAS decline of 18–31% after AI Max auto-enabled in March
• Apparel DTC brands with AOV above $200 reporting ROAS improvements of 22–34% over the same period
• Home goods merchants reporting the highest variance — outcomes dependent almost entirely on feed optimization depth
• Amazon third-party sellers running Google Shopping as a supplemental channel reporting mixed results, with profitability hinging on whether Google traffic lands on owned DTC sites or Amazon listings

How are agencies and in-house teams adapting their campaign structures?

The operational response across agencies is converging on three tactical shifts: feed-first investment, tighter asset group segmentation by margin rather than category, and aggressive use of brand exclusion lists to prevent AI Max from cannibalizing cheaper branded search traffic.

Ryan Kovacs, VP of growth at Chicago-based performance agency Decimal Digital, says his team now treats the Google Merchant Center feed as a primary creative asset — the same way they’d treat ad creative on Meta.

“We’re writing product titles the way we’d write ad headlines. We’re A/B testing descriptions through supplemental feeds. We’re using custom labels to tell Google which products have 60-point margins versus 20-point margins so the algorithm isn’t optimizing clicks on our worst SKUs. Feed management used to be a back-office task. Now it’s where the campaign actually lives.”

Kovacs notes that his team is running DataFeedWatch alongside Feedonomics for mid-market accounts — DataFeedWatch handling real-time inventory and price sync, Feedonomics managing title and description optimization — a dual-platform setup that adds roughly $800 per month per account but has become standard practice for any client spending above $30,000 monthly on Shopping.

On the budget allocation side, several agency leaders told Ecommerce Times they are reallocating spend from Standard Shopping campaigns — which are increasingly being sunset in favor of PMax — into tightly controlled asset groups within AI Max, using URL expansion controls to restrict which landing pages Google can route traffic to. The URL expansion controls, introduced alongside AI Max, have been one of the most operationally useful levers for merchants with large catalogs.

What does AI Max mean for attribution and blended CAC reporting?

Attribution is where AI Max creates its most significant downstream headache. Because the campaign type aggressively uses data-driven attribution models and pulls in YouTube, Display, and Discover placements alongside Shopping, blended ROAS numbers reported inside Google Ads frequently diverge from what merchants see in Shopify Analytics, Triple Whale, or Northbeam.

Delacroix at Ridgeline Supply Co. says she relies on Northbeam’s media mix modeling output rather than Google’s in-platform reporting to make budget decisions.

“Google will tell you AI Max drove a 6x ROAS. Northbeam tells you a different story because it can see the assist touches. The truth is somewhere in between, but I’d rather optimize to the conservative number and not kid myself about what’s actually profitable.”

This attribution gap has practical consequences for CAC management. Merchants who optimize to Google’s reported ROAS without cross-referencing a third-party attribution tool risk overspending on channels that are primarily picking up credit for purchases driven by other touchpoints — particularly email and SMS sequences, which remain heavily underweighted in Google’s data-driven model.

• Triple Whale’s Sonar attribution tool now integrates directly with PMax campaign-level data for accounts spending above $50K/month on Google
• Northbeam added a PMax-specific channel breakdown in its Q1 2026 update, allowing merchants to isolate Shopping placements from Display and YouTube within a single AI Max campaign
• Rockerbox continues to be the preferred attribution layer for merchants running heavy Meta plus Google combinations, with several agency sources citing its cross-channel deduplication as more reliable than either platform’s native reporting

Should merchants pause AI Max or lean further in?

The consensus among the agency leaders and merchants Ecommerce Times spoke with is neither: the practical answer is to run AI Max with deliberate guardrails rather than accepting Google’s default settings or abandoning the campaign type entirely.

The guardrails that are generating the most consistent results among early adopters include restricting audience expansion to in-market and affinity audiences that match first-party customer data, enabling brand safety exclusions at the account level, setting negative URL rules to prevent traffic from landing on clearance or out-of-stock pages, and using the new search-term insights report — available in AI Max accounts — to identify and exclude low-intent query themes on a weekly basis.

Pringle at Grove Commerce offers a blunt bottom-line assessment for operators still sitting on the fence.

“Standard Shopping is functionally dead. Google is going to keep pushing the machine learning controls further into the campaign, not pulling them back. The merchants who figure out how to steer AI Max — through feed quality, asset discipline, and smart exclusions — are going to have a structural cost advantage over everyone who’s still fighting the algorithm instead of working with it. The window to build that competency before it becomes table stakes is closing.”

For DTC founders managing Google Shopping in-house without agency support, the most accessible near-term action is a feed audit. Tools like Feedonomics offer a free feed health report, and Google’s own Merchant Center diagnostics dashboard flags the specific attribute gaps — missing GTINs, thin descriptions, inconsistent product type taxonomies — that most directly limit AI Max’s ability to match against high-intent queries.

The broader stakes are significant. Google Shopping remains the highest-intent paid channel available to most ecommerce operators, and for merchants with AOV above $150, it typically accounts for 35–55% of total paid acquisition volume. Getting AI Max right isn’t an optimization exercise — for most DTC brands, it’s a material factor in whether 2026 unit economics hold up.