Aisha Amaira stands at the intersection of marketing and deep technology, bringing a specialized focus to how MarTech ecosystems can be harnessed for superior customer insights. With an extensive background in CRM technology and customer data platforms, she has spent her career helping businesses bridge the gap between siloed data and actionable marketing strategies. In an era where product visibility is increasingly dictated by machine-readable data, Aisha advocates for a holistic approach to ecommerce, where technical infrastructure and search intent converge to create seamless shopping experiences.
This discussion explores the evolving role of the product feed, moving it from a purely paid advertising tool to a core pillar of modern SEO and agentic commerce. We delve into the complexities of the “unholy trinity”—the Merchant Center, on-page structured data, and the website itself—and examine how discrepancies between these layers can lead to mass product disapprovals. Aisha breaks down why SEO teams must move beyond simple on-page optimizations to co-own the feed architecture, especially as Google’s Shopping Graph grows to over 50 billion listings and AI-driven discovery becomes the new standard for consumer behavior.
Google processes product data through three distinct layers: Merchant Center feeds, on-page structured data, and the website itself. How do these systems interact, and why does their misalignment create such significant friction for ecommerce brands?
The interaction between these three layers is essentially a continuous verification loop that Google uses to establish the “truth” of a product listing. On one hand, you have the Product Feeds within the Google Merchant Center (GMC) and the Manufacturer Center, which act as a manual push of core attributes like titles, GTINs, and prices. Parallel to this is the on-page structured data, typically JSON-LD, which serves as a point of feed verification and powers those eye-catching rich results in the SERPs. Finally, there is the rendered website that both humans and bots crawl to ensure what is being promised in the data matches the reality of the user experience. Friction arises because these systems are often managed by different teams—PPC for the feed, SEO for the schema, and Dev for the site—leading to a “broken telephone” effect. When the schema says one thing and the feed says another, Google is forced to make a judgment call, and in a landscape where accuracy is the baseline for being discoverable, that call usually results in a product being suppressed or disapproved.
You’ve mentioned that price mismatches are no longer just a minor oversight but a major trigger for product disapprovals. Could you walk us through a specific instance where this data disharmony led to a visible business impact?
We recently navigated a stressful situation with an office furniture client where their products were being disapproved en masse across the Google Merchant Center. When we looked under the hood, we found a staggering level of price chaos: the website listed a chair at £34.80, the main feed matched that price, but the Merchant Center was inexplicably seeing £33.54. To make matters worse, the on-page JSON-LD schema was outputting a fourth price of £29 because it was pulling the ex-VAT price rather than the final inc-VAT figure. Because Google uses schema to verify and sometimes automatically overwrite feed prices, the system got confused by the £29 figure and the priceValidUntil field, leading to a total breakdown in trust. It was a clear reminder that if your technical layers are contradicting one another, the algorithm will pull your products from the shelf to protect the consumer experience, resulting in immediate revenue loss.
Beyond price, how do technical attributes like availability and product variants create a gap between what the PPC team manages in a feed and what the SEO team builds in schema?
The gap between feed management and schema architecture is perhaps most visible when you look at how Google handles product variants through item_group_id versus the ProductGroup schema. In a GMC feed, variants are often submitted as a simple flat list, whereas on-page schema requires a much more sophisticated, nested parent-child relationship using properties like hasVariant and variesBy. If the PPC team is optimizing a flat feed in isolation while the SEO team is building nested schema without consulting them, the mapping inevitably breaks down, making it impossible for Google to reconcile the two. We also see this with availability status; while GMC accepts four standard values—In_stock, Out_of_stock, Preorder, and Backorder—schema.org uses a different vocabulary entirely, such as https://schema.org/InStock. Without a unified strategy, you end up with “ghost products” that might be technically approved in one system but are invisible or incorrectly labeled in the organic search results.
Infrastructure failures can often be the most alarming because they happen behind the scenes. Can you describe a scenario where a site’s security settings inadvertently crippled its product visibility?
There was a particularly jarring case where a client’s products moved from fully approved to completely disapproved almost overnight, despite the feed, schema, and website content being perfectly aligned. The culprit wasn’t a data error, but a configuration change to the client’s CDN security settings that began treating Googlebot as a malicious threat. These bot protection rules were so aggressive that Google’s crawler couldn’t access the website layer to verify the data it was receiving from the Merchant Center. Since the verification bridge was broken, Google essentially decided it could no longer trust the listings and pulled them from the Shopping Graph entirely. This illustrates why you need a “systems thinker” who looks across crawl behavior, infrastructure, and feed health simultaneously; a PPC manager might see the disapproval, but they wouldn’t necessarily think to check the CDN’s firewall logs.
Traditionally, the Merchant Center has been seen as a PPC tool. Why is it now essential for SEO teams to step into the room and treat the feed as a core search asset?
The logic that the Merchant Center is only for Paid Shopping is fundamentally outdated because the feed now serves as the primary infrastructure for your entire product presence, including free listings and AI responses. When SEOs stay out of the conversation, the feed often becomes a generic data dump from a platform plugin, filled with thin attributes and titles that were never built with search intent in mind. PPC teams often try to patch this with feed rules or supplemental feeds, but they are usually optimizing for bidding efficiency rather than the broad taxonomy and keyword depth that an SEO instinctively understands. By co-owning the feed, SEOs can ensure that product titles are front-loaded with high-intent keywords and that the taxonomy is refined so products aren’t buried in generalized categories. It’s about transforming the feed from a cold database file into a commercially vibrant asset that speaks the same language as the customers’ queries.
Google’s Shopping Graph now contains over 50 billion listings and powers experiences like AI Overviews and Gemini. How does feed quality influence a merchant’s overall “trustworthiness” in this massive ecosystem?
Feed quality has evolved into a holistic signal that Google uses to evaluate a merchant’s reliability across all its surfaces, not just for ad eligibility. Through programs like the Shop Quality program, Google measures signals such as approval rates, shipping data completeness, and the clarity of return policies to assign rankings and badges like the Top Quality Store icon. If your feed is riddled with attribute gaps or recurring disapprovals, it sends a signal to Google that your data is unreliable at scale, which can diminish your visibility in both paid and organic grids. In the context of the Shopping Graph, consistency is the currency of trust; when Google can verify your data across the Merchant Center, your landing pages, and your schema, it feels confident surfacing your products in high-stakes environments like AI Overviews. Essentially, a healthy feed is no longer just a campaign requirement—it is a competitive factor that determines whether you are invited to the “AI table” at all.
The concept of “Agentic Commerce” suggests that discovery is moving away from being purely human-led. What technical hurdles must brands overcome to ensure an AI agent can actually find and purchase their products?
As we move toward agentic commerce, the definition of “discoverable” shifts from having a pretty website to having a robust, machine-readable representation of your entire catalog. AI-powered surfaces like Gemini and AI Mode draw heavily from the Merchant Center data to answer commercial queries, meaning thin attributes will lead to a significant disadvantage during the discovery phase. But even if an agent finds your product, the “purchasability” phase is a whole different technical challenge involving the raw HTML and the accessibility tree of your site. If your site uses “div soup” instead of semantic buttons or has a layout that shifts during loading, a machine agent won’t be able to interpret the “Add to Cart” function, and the transaction will fail instantly. Brands must start treating their site’s accessibility tree and conversational commerce attributes—like compatibility and substitutes—as critical infrastructure for the future of automated shopping.
Implementing shared ownership between SEO, PPC, and Dev is notoriously difficult in large organizations. What practical steps can teams take to move toward a unified “Product Truth” layer?
The first step is to break the habit of only talking when something breaks and instead build a cross-departmental monitoring layer that alerts both SEO and PPC teams to disapproval spikes simultaneously. We advocate for a monthly structured audit where teams sit down together to compare feed attribute completeness against on-page schema and website reality, asking if prices match and if required attributes are present across all three layers. It’s also vital to document a single “source of truth”—whether that’s the CMS, the ERP, or the feed platform—so everyone knows which system is authoritative for specific attributes. Finally, SEOs need to be present at the table during feed architecture decisions, not to veto the PPC team, but to ensure that custom labels and taxonomy choices are made with a shared understanding of how Google’s algorithm reconciles data. When these departments act as a unified front, you create a product presence that is far more resilient than the sum of its parts.
What is your forecast for the role of the Merchant Center in the broader SEO landscape over the next three years?
In the next three years, I anticipate the Merchant Center will become the central nervous system of ecommerce SEO, effectively merging the boundaries between the “technical” and “commercial” web. We will see Google move toward a more integrated processing system that may eventually unify schema markup and feed data into a single standard to eliminate the friction of the current “unholy trinity.” As AI agents become the primary shoppers, the importance of the “Product Truth” layer will skyrocket, making the Manufacturer Center and detailed attribute mapping the most important levers for organic growth. SEO will no longer be about just ranking a URL, but about managing the integrity of a product’s digital twin across a fragmented ecosystem of AI modes, visual search grids, and automated checkout systems. The brands that stop treating their feed as a secondary PPC task and start treating it as their most valuable data asset will be the ones that dominate the Shopping Graph.