Google's framing of AI Max for Shopping is deliberate. It is not positioned as a campaign update or a feature refresh. It is positioned as something built for a different kind of search entirely - one where shoppers do not arrive with a finished query, but where discovery and intent form simultaneously. That framing matters more than any individual feature within it.
For retailers and their agencies, this represents a real shift in how to think about Shopping campaign strategy. The question is no longer just 'how do I show up for shopping queries?' It is 'how do I show up at the moment someone starts becoming a shopper?'
The Gap Between Traditional Shopping and Modern Search Behaviour
Standard Shopping campaigns were built around a relatively simple model. A user knows what they want, types it in, sees product listings, clicks. The campaign structure reflected this - product feeds, match types, bidding against known queries. It worked because search behaviour was largely transactional and fairly predictable.
That model has been cracking for a while. Conversational search, AI Overviews, and multi-step browsing sessions have all changed how people move from vague interest to purchase intent. Someone searching 'what should I wear to a winter wedding as a guest' is a potential clothing buyer. Traditional Shopping campaigns would never have reached them at that moment. They were structurally blind to it.
AI Max for Shopping is Google's explicit acknowledgement that this gap exists and that closing it requires a fundamentally different approach - one where the AI interprets context and intent rather than waiting for a clean, categorical query to match against.
What This Means for Feed Quality and Product Data
Any campaign that relies on AI-driven matching is only as good as the data it works with. This is where many retailers will find the real work sitting. If your product feed was built to satisfy a rigid match model - titles stuffed with exact keywords, minimal descriptive detail, attributes filled in at the bare minimum - it will underperform in a system trying to interpret broader context.
Richer product data is the foundation. That means detailed, accurate product titles that describe the item clearly rather than just target a keyword. It means complete attribute sets - materials, use cases, compatible contexts, sizing information where relevant. It means descriptions that genuinely communicate what the product is and who it is for, rather than a line of keyword repetition.
For retailers managing large catalogues, this is not a quick job. But it is the single most important lever available. An AI-driven campaign that cannot understand your products cannot match them intelligently. The feed is the brief you give the system - and a poor brief produces poor results regardless of how sophisticated the AI is.
The Attribution Challenge That Comes With Discovery-Led Campaigns
When a campaign targets users at the discovery stage rather than at the point of purchase intent, the conversion path gets longer. Someone who sees a Shopping ad during a browsing or discovery session may not convert for days, or may convert through a different channel entirely. Last-click attribution models will misread this badly.
Retailers running AI Max for Shopping alongside more intent-focused campaigns need to think carefully about how they are measuring each. Data-driven attribution through Google Ads is a better fit here than position-based models, because it accounts for partial credit across a longer path. But even that has limits if your conversion window is set too short or your cross-channel tracking has gaps.
This is a practical concern, not a theoretical one. If discovery-led campaigns are measured purely on last-click ROAS, they will consistently look weak compared to campaigns targeting high-intent queries - and budget decisions will be made accordingly. The result is that you optimise away from upper-funnel activity and then wonder why your conversion volume starts declining further down the line.
Smart Bidding Configuration in a Discovery Context
Smart Bidding strategies need recalibrating when the campaign goal is reaching shoppers earlier in their journey. Target ROAS is a reasonable choice for campaigns where your data set is strong and your margin structures are well defined. But set a Target ROAS that was built on conversion data from intent-heavy campaigns, and you are likely setting an expectation the discovery-stage campaign cannot meet in the short term.
Maximise Conversion Value with a ROAS target set more loosely - or Maximise Conversions during the learning phase - gives the system more room to operate. The aim during that period is to generate enough conversion data across the broader funnel to let Smart Bidding calibrate properly. Cutting the campaign or tightening the target before that data exists is one of the most common ways these campaigns get strangled before they can perform.
Portfolio bidding strategies, where you group campaigns and let the bidding algorithm optimise across a shared target rather than per campaign, are worth considering for retailers running multiple Shopping campaign types simultaneously. They can smooth out performance variance between discovery and conversion-stage campaigns rather than treating each in isolation.
The Broader Signal: Search and Shopping Are Converging
AI Max for Shopping is part of a consistent direction across Google Ads over the past two years. The boundaries between campaign types are softening. Performance Max already spans search, shopping, display, YouTube, and Discover. Demand Gen sits across YouTube and social surfaces. Now Shopping is explicitly extending into the discovery and conversational search space.
For retailers, this means the old approach of managing Shopping campaigns as a discrete, self-contained unit - separate from brand, separate from broad match search, separate from video - becomes harder to sustain. The platforms are increasingly integrated, and the AI systems connecting them work better when campaigns are structured to complement each other rather than compete.
That is not a reason to abandon campaign-level thinking entirely. Granularity still matters for control, reporting, and budget allocation. But the strategic frame needs to shift. Shopping is no longer just a channel for capturing demand. It is becoming a channel for creating it. Retailers who build their campaigns - and their measurement frameworks - around that reality will be better placed than those still optimising for a search model that is already changing beneath them.