AEO

April Recap

Retail Search Intelligence

No. 2 Google's Algorithm Update Trifecta, AI Search, and Agentic Shoppers, oh my!

True Story. I was shopping for a new SPF sunscreen for my face for the summer (but also, how is it May??). So I started my research in ChatGPT, and typed: "What is the best lightweight, noncomedogenic sunscreen with SPF30+ for sensitive skin in the summer. Needs to be sweat-proof and not leave any residue."

That search is 24 words long.

Here's what's interesting: the brands that showed up weren't necessarily the biggest ad spenders or the highest-ranking SEO pages. They were the brands whose product content was structured enough for an AI to actually understand, cite, and confidently recommend. That's the game now.

If your product page just says "SPF 30, 1 oz, $18.99," the AI skips you. If it explains what makes your formula ideal for active summer days, how it layers under makeup, and what real customers with sensitive skin have experienced? You get the recommendation.

Spoiler alert: attributes like these and others including google's new conversational attributes can be updated across your entire catalog automatically to ensure your products are being surfaced when shoppers search in AI mode, ChatGPT, etc.

What this means for you:

UCP runs entirely on structured data quality. If your Google Merchant Center feed has incomplete attributes, stale pricing, or missing variant-level data, AI agents won't be able to retrieve your products at query time. You're not "penalized" in the traditional search sense, you're simply invisible. Your competitors with clean, attribute-rich catalogs get the AI recommendation instead.

NOTE: Shoppers arriving via AI search citations convert at 9× the rate of traditional organic visitors. Is your product data ready??

Industry Pulse

Google Fired Three Algorithm Shots in Six Weeks, And They've All Now Landed - March was the busiest algorithm month Google has had in years. Three separate updates rolled in rapid succession, and they've all completed. Here's the recap every retailer needs:

1. The March 2026 Spam Update (March 24–25) — Completed in under 20 hours, making it the fastest spam update in Google's documented history. It targeted parasite SEO, manipulative link schemes, mass AI-generated content without human oversight, and cloaking. Google didn't introduce new rules — it just got dramatically better at enforcing the existing ones.

2. The February 2026 Discover Update (Feb 5–27) — The first-ever core update Google scoped exclusively to Google Discover. It targeted clickbait, rewarded original reporting, and narrowed which domains appear in US Discover feeds. Currently English-only, with global expansion confirmed later in 2026.

3. The March 2026 Broad Core Update (March 27–April 8) — The first broad core update of 2026, completing in 12 days. It affected all sites, all languages, all regions globally. SEMrush Sensor hit 9.5/10 volatility — among the highest ever recorded. Google described it as better surfacing "relevant, satisfying content" which in plain English means: thin content lost, original content won. Rankings have supposedly now settled, but worth checking!

What you need to know:

The spam and core updates ran back-to-back intentionally. Sites the spam update weakened entered the core update already damaged. The combined hit was harder than either alone. If your traffic moved in late March, don't diagnose it through one lens. You may be seeing two or three separate events layered on top of each other. Open Search Console: if your drop started March 24–25, it's spam-related; if it started March 27+, it's core. The fixes are completely different.

Google provided a list of questions to consider if your site is hit by a core update.

Live Case Study: Gap Inc. and Hibbett Just Became Shoppable Inside AI

At Shoptalk on March 24, Gap Inc. announced it is one of the first major apparel brands to go live with Google's Universal Commerce Protocol (UCP). Old Navy, Gap, Banana Republic, and Athleta are now discoverable, comparable, and purchasable directly inside Google's AI Mode and the Gemini app.

The key detail: Gap remains the merchant of record. They keep the customer relationship, the loyalty data, and the sale. The AI is the interface, not the intermediary.

"We are not pursuing AI for novelty. These partnerships are about solving real customer problems." — Sven Gerjets, CTO, Gap Inc.

UCP now supports real-time catalog access, multi-item carts, identity linking (so loyalty members get their pricing automatically), and direct offers. Best Buy and Wayfair are also in. The infrastructure is being built right now, and Google is simplifying UCP onboarding in Merchant Center to accelerate adoption.

Hibbett is also one of the first retailers to officially launch an "AI store" directly in ChatGPT. Think about all of the retail apps on your phone (I have Target, Sephora, etc.). But that limits me to see only specific products with specific searches within a rigid taxonomy. Now shoppers can have full on conversations just like they would in ChatGPT, but within the Hibbett catalog. So I can say "I need new shoes for an outdoor conference in Miami next week that looks good with a dress/skirt" >> "I also have wide feet and need support because I have knee problems" (true story).

The AI Surfaces Measurement Problem: What to Track?

AI search is driving real brand exposure, but it's almost impossible to measure with existing tools. The reason: cited sources in Google AI Mode and ChatGPT change month-to-month at a 40–60% rate. A keyword ranking holds for months. An AI citation can disappear in weeks. And unlike Google, there's no place to UI to see your "stats" or analytics.

What you can start tracking this month:

AI Answer Mentions: Track which products or pages appear in generative responses across ChatGPT, Perplexity, Gemini, and Google AI Overviews.
Prompt Volume: Understand the natural language questions your category generates and audit whether your content answers them directly.
Schema Health: Monitor error-free structured data coverage across all high-revenue pages. Broken schema = invisible to agents.
AI Surface Share: Measure presence and position in AI shopping panels for target queries — this is the new "share of voice" metric for retail.
Assisted Revenue: Track how AI-surfaced pages influence conversion, AOV, and revenue across the funnel, not just as a last-click attribution.

NOTE: We offer this in our Optiversal AI Visibility Dashboard

Product Spotlight: Enriched conversational attributes across platforms

Google recently launched its new conversational attributes which are designed to help AI agents understand and recommend products within conversational, "agentic" commerce. This includes Google's new AI Shopping Agent.

Search is becoming a conversation, and we want to ensure your products speak the language for both humans and AI. So we’ve expanded our ability to enrich conversational attributes across all SKUs for all major platforms:

Meta: Added support for dozens of rich, social-first attributes.
OpenAI: Our feeds now support Q&A structures to power discovery within ChatGPT and other LLMs.
Google: Now supporting conversational attributes in Beta.

What we’re capturing (examples):

FAQs: (e.g., "What makes this laptop good for college?")
Occasions: (e.g., Back to School, Graduation)
Technical Specs: (e.g., 256GB RAM, 17” display)
Contextual Data: Style, Keywords, Themes, and more.

ICYMI - Check out the “Personas" in Settings

Personas have been introduced into the Brand tab, allowing you to enter your own identified customer personas. You are likely already using these personas for email, ads, and marketing campaigns, but now you can create persona-specific copy enrichment for where the sale actually happens - on the product detail page (PDP).