Google has been steadily adding AI features to Merchant Centre throughout 2026 - first the Merchant Advisor assistant, then AI performance reports, and now AI summary insights, search suggestions and persistent chat history directly on the dashboard. Taken individually, each addition looks incremental. Together, they represent a deliberate shift in how Google expects advertisers and feed managers to interact with product data.
For anyone running Performance Max campaigns with Shopping inventory, this matters more than it might first appear. Merchant Centre has always been the engine room for Shopping - the place where product data quality ultimately determines what gets served, to whom, and at what cost. Wrapping AI summaries and insight prompts around that data changes how you interrogate it, and potentially changes what you act on.
What the AI Features Actually Do
The new AI summary insights appear to surface digest-style overviews of account and feed health directly on the Merchant Centre dashboard. Rather than drilling into individual reports, advertisers get a headline view of what the system considers significant. The addition of search suggestions - effectively prompt recommendations for querying your own data - sits alongside previous chat history, suggesting Google is positioning the Merchant Advisor as an ongoing conversational tool rather than a one-off diagnostic.
The AI performance reports that arrived earlier in 2026 are the more commercially relevant piece. Performance reporting inside Merchant Centre has historically been thin compared to what you can pull from Google Ads directly. If these reports are surfacing product-level performance signals with enough granularity - impressions, click-through rate, conversion contribution - they could genuinely close a gap that has made feed-level optimisation harder than it should be.
The honest caveat is that the quality of any AI summary is only as good as the data it is summarising. Merchant Centre has its own data freshness and attribution limitations. Treating an AI-generated digest as a substitute for proper product-level analysis in Google Ads would be a mistake.
Why Feed Quality Remains the Actual Lever
There is a risk with any AI-assisted insight layer that it creates the impression of being on top of something without requiring the underlying discipline. Merchant Centre AI summaries might tell you that a product category is underperforming - but the root cause is almost always in the feed itself. Weak titles, missing attributes, incorrect GTINs, poor image quality, or price competitiveness issues are not problems an insight summary fixes. They are problems that require manual feed work.
For Performance Max campaigns in particular, the feed is doing a significant amount of the targeting and creative work that advertisers used to control through keywords and ad copy. Google's systems use product titles, descriptions, and category attributes to determine query matching and audience relevance. An AI insight that says impressions are down on a product group is useful as a prompt - but the investigation still needs to happen at the attribute level.
Practically, this means the value of Merchant Centre AI features is as a triage tool, not a replacement for structured feed auditing. Use them to identify which product groups warrant attention, then go deeper using supplemental feeds, feed rules, or a dedicated feed management process to address the underlying issues.
The Attribution Problem Has Not Gone Away
One of the persistent frustrations with Shopping performance analysis is that attribution at the product level is genuinely difficult. Google Ads reports on campaign and ad group performance. Merchant Centre reports on feed health and impressions. Tying product-level contribution to actual revenue - particularly for advertisers with large catalogues - requires either custom reporting in GA4 or Looker Studio, or pulling data from both platforms and reconciling it manually.
The AI performance reports in Merchant Centre could help here if they are surfacing conversion-related data alongside impression and click metrics. But this is worth verifying rather than assuming. The risk of acting on incomplete performance data is that you deprioritise or pause products based on Merchant Centre signals that do not account for assisted conversions or cross-device behaviour that only shows up in GA4 with proper ecommerce tracking in place.
If you are running Shopping campaigns at any meaningful scale, the priority should be ensuring your GA4 ecommerce implementation is solid - item-level purchase events, correct item IDs that match your Merchant Centre feed, and revenue reporting that you can cross-reference against what Google Ads and Merchant Centre are showing. AI summaries in Merchant Centre are a useful signal layer on top of that foundation. They are not the foundation itself.
How Agencies and In-House Teams Should Use This
For agencies managing Shopping accounts on behalf of clients, the Merchant Advisor chat history feature has a practical benefit beyond the AI insights themselves. Having a persistent log of queries and responses inside Merchant Centre creates a lightweight audit trail of diagnostic activity - useful for demonstrating the work that goes into feed management, which can otherwise be invisible to clients who only see campaign performance numbers.
For in-house teams, the search suggestions feature - where the AI recommends what to query - is worth treating sceptically until you understand what it is actually prioritising. If it surfaces the same handful of obvious prompts regardless of account specifics, it is not adding much beyond a nudge to look at things you probably already knew to check. The real test is whether it identifies non-obvious issues: products with high impressions and low conversion rates, seasonal catalogue items that need attribute updates, or feed errors that are suppressing specific SKUs without surfacing as formal disapprovals.
The broader pattern here - Google adding AI-assisted insight layers across its advertising tools - puts more emphasis on advertisers being clear about what decisions they are willing to delegate to automated interpretation and what requires human judgement. Merchant Centre AI features are at the informational end of that spectrum. They are surfacing summaries and suggesting questions. The decisions about feed investment, catalogue prioritisation, and campaign structure remain yours to make.
What to Actually Do Next
If you have not explored the AI features in Merchant Centre recently, it is worth a proper look - not to take the summaries at face value, but to understand what signals the platform considers significant. That tells you something about how Google's systems are interpreting your feed, which is useful context for Performance Max campaigns where you have limited visibility into how product data is being used for targeting.
Audit your feed quality independently of whatever Merchant Centre surfaces. Check title structure against how users actually search for your products. Confirm your GTINs are accurate. Review your custom labels to make sure they are still serving a purpose in your bidding and segmentation strategy. These are the inputs that determine Shopping performance - the AI insights in Merchant Centre are, at best, a prompt to do that work more regularly.
And make sure your measurement is in order before you draw conclusions from any new reporting layer. If your GA4 ecommerce tracking is incomplete, or your Merchant Centre item IDs do not match your analytics implementation, you are working with a distorted picture regardless of how the AI chooses to summarise it.