New AI Performance Insights: Navigating the Merchant Center Landscape
Google's recent foray into AI performance reporting within Merchant Center has provided retailers with a new, if limited, lens through which to view their product visibility. As the digital shopping experience shifts from traditional, search-query-based interaction to AI-driven, conversational interfaces—like AI Overviews and AI Mode—understanding how your products appear in these results is becoming a cornerstone of visibility strategy. However, as the initial enthusiasm for this release balances against its enduring limitations, it is clear that we are in a necessary transition period. It is a period where the data we do have offers genuine value, but the granular metrics we actually need to optimize for ROI remain elusive.
What Merchant Center’s AI Performance Insights Actually Offers
The reporting package, initially unveiled at Google Marketing Live, provides a glimpse into the AI discovery surfaces for your product feed. You can access it under Analytics -> Products -> AI performance.
While it does not provide granular query-level data, it does offer a thematic, conceptual view of customer intent. The metrics are grouped into a few key areas:
- Query Type: Where Google categorizes the underlying intent of the search—are they researching specs, looking for reviews, or searching by category?
- Query Frequency: A measure of how popular a concept is in the emerging AI-driven shopping journey.
- Phase of Shopping Journey: An indicator of where the user is when they ask the question.
- Product Terms: The specific language shoppers use to describe what they want—think “arch support” or “maximum cushioning” instead of just “running shoes.”
- Share of Voice (SOV): How your AI impressions compare to your competitors in the set defined by Merchant Center.
This allows for a strategic pivot: rather than chasing "keyword gold" as we did in the search bar era, retailers must now focus on "intent fulfillment." The queries and terms surfaced here are input triggers to improve your feed, not a list of target queries. By understanding if shoppers are searching based on feature comparison or aesthetic appeal, you can tailor your product data to better align with these AI-generated journeys.
Strategy: Translating Insights into Product Action
The real value of this report lies in bridging the gap between aggregate data and product-level optimization. Here is how you can practically apply these insights:
- Audit Your Vocabulary: If the report highlights that shoppers use specific terms to ask about your product category—for instance, "performance breathable tech-fabric" rather than simply "athletic shirt"—proactively integrate this vocabulary into your feed.
- Address Intent Mismatches: If you see shoppers inquiring about product features in AI queries that your current feed doesn't highlight, this is a clear signal to refine your product data. Ensure those attributes are explicitly mapped in your feed fields.
- Competitive Benchmarking: While the Share of Voice (SOV) metric is imperfect, use it to benchmark your visibility against competitors within the set predefined by Google. If your SOV consistently drops for specific query types, investigate those items for deeper competitive feed-quality issues.
Why These Aren’t Traditional SEO Metrics
It is vital to maintain perspective: these numbers do not equal traditional SEO volume or click-through behavior. Indeed, they represent a fundamentally different data paradigm.
- The Organic-Only Constraint: Any paid ad activity is entirely excluded from this report, which simplifies the dataset but also renders it incomplete if you are trying to measure a holistic presence.
- The SOV Limitation: Share of Voice in this report is relative and volatile. Since Google defines the competitor set, it lacks the flexibility for manual competitive analysis and can be highly biased by your total impression volume. If you have low impression volume, the report shows zero. If you have no competitors listed, it might show 100%. Don’t treat these numbers as absolute truth.
The Broader Data Gap: Beyond the Product Feed
The fundamental concern for the industry is what simply isn't here: click-through data and individual, exact-match query insights. This is a recurring theme—the "black box" of AI reporting from Google.
For editorial and affiliate publishers, who do not have a product feed, the situation is even more opaque. They remain in an analytical vacuum, awaiting more robust reporting from Google Search Console. The experimental generative AI reports in Search Console for a subset of UK sites indicate that we are eventually intended to have visibility here, but the granular data needed for actionable SEO remains absent.
The Push for Transparency: Regulators and the Industry
The industry is not sitting still. Organizations and regulators, including the UK’s Competition and Markets Authority (CMA), are pressuring Google to provide more transparency. They are specifically calling for clarity on impressions, click-through rates, and separated generative AI traffic to ensure a level playing field for publishers and retailers.
This push is essential, as the current level of reporting creates a significant imbalance. Without this data, retailers and publishers are operating on intuition rather than actionable intelligence when it comes to the AI-driven search environment. The current state is, in many ways, an attempt by Google to curate the AI experience while keeping the data silos tightly controlled.
Moving Toward an AI-First Visibility Framework
The era of simple keyword tracking is fading. As we move forward, our strategy must evolve towards an AI-first framework that prioritizes "intent-centric" optimization. This means:
- Context-Driven Content: Providing enough context for LLMs to interpret your product or content accurately.
- Structured Data Prowess: Enhancing structured data to make your offerings more readily accessible by AI agents.
- Adaptive Monitoring: Treating these AI reports as the evolving narrative they are, rather than static performance goals.
Ultimately, the crumbs of data provided by Google's current AI reporting are better than nothing, but they serve as a reminder that understanding our presence in a world of generative AI is a marathon, not a sprint. We must continue to push for the transparency that is necessary for a healthy digital ecosystem.
[Source: https://www.searchenginejournal.com/googles-ai-search-data-is-growing-but-the-gaps-remain/582558/]
