AI Visibility vs Ecommerce SEO: What Changes When You Already Have an Agency

You already have an SEO agency. You already pay for Semrush. Rankings are fine. Then someone on the leadership team asks whether the brand shows up when shoppers ask ChatGPT or Google AI Mode what to buy, and nobody in the room knows. The honest follow-up question is the one this post answers: what is actually different about AI visibility, and what does your team have to do that the agency is not already doing?

The short version. AI visibility and ecommerce SEO share one foundation, which is a crawlable site with complete product information and an eligible Merchant Center feed. AI visibility adds a different question set, a different result format and a measurement your SEO stack does not produce. It does not replace SEO. An earlier version of this post said traditional SEO was irrelevant to AI. That was wrong, and the rest of this explains why.

What carries over

Google is explicit that AI Overviews and AI Mode draw on Search. Its documentation for site owners says the same foundational SEO practices apply, that a page must be indexed and eligible for a snippet to appear as a supporting link, and that there are no additional technical requirements. Crawlability, structured data, clean titles and fast pages serve both the blue links and the AI answer.

ChatGPT shopping is built from the same raw material: your product pages, your Google Merchant Center feed and what third parties say about you. If your agency has kept the site crawlable and your feed approved, that work is already contributing to AI visibility. The ecommerce SEO work we do for brands starts from that shared base.

What is different

The question is longer and the answer is shorter

A keyword is "running shoes flat feet". A prompt is "I have flat feet and I am training for a half marathon, what running shoes should I look at under $150?". Google documents that AI Mode and AI Overviews use a query fan-out technique, running several related searches across subtopics before composing a single response. The response names a handful of products. There is no page two.

Ranking is an input, not the result

Surfer SEO analyzed 173,902 URLs and found that about 68% of pages cited in Google AI Overviews did not rank in the top ten for the main query or any of its fan-out queries. Ranking for the main query and its sub-questions helps, and the same study shows pages that rank for several fan-outs are cited far more often. It is not sufficient on its own. The assistant is choosing passages that answer the sub-questions, wherever they live.

Non-branded prompts decide more

Shoppers who already know you search your name and find you through normal search. Shoppers who are still deciding describe a need, and that is where an assistant chooses among brands. We laid out the difference in branded versus non-branded AI visibility. Your SEO reporting is usually dominated by branded traffic, which is why AI visibility can be poor while the SEO dashboard looks healthy.

The volume is small and growing

Bain's analysis of Sensor Tower data found that shopping prompts rose from 7.8% to 9.8% of all ChatGPT searches between January and June 2025, on top of roughly 70% growth in total prompts over the same period. For most stores, AI-referred sessions are still a small share of traffic. The reason to care is the kind of session: a shopper arriving from an answer that already recommended you.

What shoppers ask that your product page does not answer

Most AI visibility gaps are product information gaps. Read your best-selling product page the way a prompt would and check whether it states the facts a buyer asks before choosing:

• Who it is for: the use case, the skill level, the body or skin type, the room or vehicle it fits.
• What it is made of and how it is measured: materials, weight, dimensions, capacity, concentration.
• How it compares: what makes it the lighter, warmer, gentler or quieter option against the obvious alternative.
• What is in the box, what it works with, and how it is cared for.
• Price, availability, shipping and returns, matching what the Merchant Center feed says.

A title written for a keyword reads like "Men's Running Shoe Lightweight Marathon Training Black/Red". A description written for a prompt says who the shoe suits, how it is cushioned, what it weighs, and how it grips on wet roads, in plain sentences. The second version answers the sub-questions. It also reads better for a human, which is why this work tends to help conversion on the page as well.

How to measure it

Semrush tells you where you rank. It does not tell you whether ChatGPT named you when a shopper asked for the best option in your category. The measurement for AI visibility is a fixed set of shopper questions, written by persona, run on a schedule across ChatGPT, Google AI Mode and Perplexity, with the results logged: named or not, product card or link, how the product was described, which competitors appeared instead.

That gives you a before and after. This is what we saw, this is what we changed in the product information, this is what shows up now. Pair it with AI referrers tagged as their own channel in GA4 or Shopify, which we walked through in how to measure AI referral traffic, and keep answer appearances, sessions and orders as three separate numbers.

What to fix first

1. Pick the products that carry revenue. A few hundred items, not the whole catalog.
2. Fix the product information on those pages: the facts above, in shoppers' words, written from your own specifications and support data. This is the product content work that every surface reads.
3. Make the Merchant Center feed eligible and current for the same products: disapprovals, missing identifiers, price and availability mismatches, categories Google guessed.
4. Set up the prompt check and the baseline, then repeat it after the fixes ship.

None of this conflicts with your agency's work. The agency owns rankings and technical health. Your team owns the catalog. AI visibility is the layer where the product information both depend on gets judged by an assistant instead of a ranking algorithm, and that is a job your SEO retainer was not scoped to do.

If your agency owns rankings and your team owns the catalog, AI visibility starts with the product data both depend on. This is the AI visibility work Paz does for lean ecommerce teams. A Paz expert finds the gaps in how your products show up in ChatGPT and Google AI Mode, you approve the fixes, the platform publishes them and tracks before and after. If you want to know whether you show up, talk to us.