By Vicky Sidler | Published 28 August 2025 at 12:00 GMT+2

Once upon a time, people typed “best bed sheets” into Google, clicked a few links, and maybe got distracted by a YouTube video of someone folding a fitted sheet perfectly.

Now? They ask ChatGPT, “I sleep hot, have eczema, and hate scratchy fabric. What’s a soft, breathable sheet that won’t irritate my skin?”

Same goal. Very different search rules.

According to Leigh McKenzie’s recent guide for ecommerce sellers, AI search is turning traditional SEO into a multi-layered, multi-platform game. If you want your products to be part of the answer when customers ask the internet what to buy, you need more than keywords.

You need what she calls relevance engineering.

Let’s break that down—plain English, practical steps, no fluff.

TL;DR

• AI search tools now recommend products based on prompts, not just keywords

• Your product pages need to be crawlable (not blocked) and readable (even without JavaScript)

• Structured schema markup helps AI understand what you sell

• Product feeds are now fuel for ChatGPT, Perplexity, and Google Gemini

• AI bots are crawling your site—you should track them

• Optimizing for personas and prompts matters more than old-school keyword lists

• Your brand must show up across the web—not just your site

• Visibility is now probabilistic—your results vary based on how people ask

Need help getting your message right? Download the 5-Minute Marketing Fix.

Table of Contents

Why AI Search Is Changing the Rules:

AI tools don’t “search” like we do. They interpret. When someone types a long, specific question, these bots scan their training data, recent web content, and product feeds—and then make a recommendation.

So if your product info isn’t showing up in the right places (or isn’t understandable to an AI), you won’t make the cut.

That’s where AI search optimization comes in.

Step 1—Make Sure Bots Can See Your Products:

First things first—your product detail pages (PDPs) need to be:

• Crawlable (bots can access them)

• Renderable (bots can read the content in plain HTML)

If you’re blocking GPTBot or PerplexityBot in your robots.txt file, they won’t crawl your site. That means your products won’t show up in AI answers. Double-check your settings.

Also, if your key info (like descriptions or pricing) loads via JavaScript, tools like ChatGPT and Perplexity may not see it. They can’t parse JavaScript—they read raw HTML. To test this, disable JavaScript in your browser and reload the page. If your product disappears, so will your visibility in AI.

Quick fix: Talk to your dev team about server-side rendering or prerendering important content.

Step 2—Add Structured Schema Markup:

Think of schema markup as a cheat sheet for AI. It tells bots what’s on your page in a format they understand instantly.

Use Schema.org’s “Product” type to describe:

• Product name

• Description

• Brand

• SKU or GTIN

• Price and availability

• Reviews and ratings

Stick to JSON-LD format, embed it in the raw HTML, and make sure it reflects what’s actually on the page.

Bonus: Add FAQ and HowTo markup if it makes sense. The more context you provide, the more likely AI tools are to feature your products in responses.

Step 3—Build a High-Quality Product Feed:

Think of your feed as the shopping list AI bots use to answer questions. The basics include:

• Title

• Description

• Price

• Product URL

• Availability

• GTIN, MPN, Brand

• Image URL

Layer in extras like:

• Product categories

• Variants (colour, size)

• Shipping details

• Review count and star rating

And write in plain English. Not “moisture-wicking upper”—say “keeps your feet dry.” Use the words your customers use in reviews and support tickets.

Then submit your feed to platforms like:

• Perplexity’s Merchant Program

• Google Merchant Center

• OpenAI (pilot program)

This makes it easier for AI tools to feature your products directly in answers—including product carousels and conversational recommendations.

Step 4—Monitor AI Bots:

Just because you opened the door doesn’t mean anyone’s visiting.

Set up server log tracking or use your CDN to monitor traffic from:

• GPTBot (ChatGPT)

• OAI-SearchBot (OpenAI search)

• PerplexityBot (Perplexity AI)

• Googlebot (various flavours)

Track:

• Which pages they visit

• How often they come back

• Whether they crawl your product feed

This helps you see what bots are actually indexing—and where you might be falling short.

Step 5—Think in Prompts and Personas:

Old SEO: Keywords like “best running shoes”

New SEO: Prompts like “I’m training for a marathon in rainy weather. What’s a lightweight shoe with good grip?”

See the difference?

Your content, feeds, and page copy need to reflect how real people talk—and shop—in natural language.

Build prompt-based content around:

• Personas (e.g. hot sleepers, allergy sufferers)

• Use cases (e.g. travel, gifts, everyday basics)

• Problems (e.g. sheets that bunch up, shoes that rub)

If you’ve ever built a StoryBrand BrandScript or Duct Tape customer persona, you already know how to do this. This is just applying it to your product descriptions and AI-facing content.

Step 6—Show Up Everywhere (Not Just Your Site):

LLMs learn from the whole internet. That means your brand mentions on Reddit, Quora, Trustpilot, YouTube, and third-party blogs all shape how AI sees you.

So even if your product pages are perfect, if no one’s talking about your brand out there—AI won’t either.

Where to get mentioned:

• Review platforms (Google, Amazon, Trustpilot)

• Forums (Reddit, Quora)

• YouTube videos (yours or creators’)

• Affiliate blogs and product roundups

Get your product into these spaces. And make sure the language used aligns with what your ideal customers would say in an AI prompt.

Step 7—Track Your AI Visibility:

Here’s the weird part. You might show up in one AI response and vanish from the next—even if the prompts are similar.

Why? Because AI results aren’t static. They’re generated on the fly based on your visibility, trust signals, and how the prompt is worded.

So stop tracking fixed keyword rankings. Start tracking prompt coverage instead.

How to do it:

• Build a list of likely prompts

• Use tools like Semrush AI Toolkit or Peec.AI to check if you show up

• Note where your brand is cited, and what tone the AI uses

• Compare yourself to competitors over time

Think of it like listening for echoes. If your brand keeps showing up in the answers, you’re doing it right.

The End Goal—Be the Answer, Not Just the Option:

This isn’t just about “optimising for AI.” It’s about making sure your products—and your message—show up wherever people are looking.

Your brand should feel like the natural answer when someone types in a long, confused, very human prompt like:

“I’ve just moved into a tiny apartment, I’m allergic to dust, and I hate ironing. What’s the best low-maintenance bedding set under $100?”

If your store doesn’t have a product page that matches that energy, someone else’s will.

Want to be that answer?

Start by getting your messaging right. The 5-Minute Marketing Fix will help you craft one sharp, clear line that cuts through the AI noise and grabs attention across every platform—bot or human.

👉 Download it free here

Related Articles

New Research Reveals How ChatGPT, Google AI, and Perplexity Source Info

If you're optimising for AI search, you need to know where these tools pull their data from. This breakdown shows how different AIs favour different platforms—and what that means for your ecommerce strategy.

AI Tools May Mention Your Brand Without Sending Clicks

If you're following the 7 steps to show up in AI results, this post explains why tracking AI mentions matters—even when they don't lead to clicks. A must-read for measuring AI visibility impact.

Some Shared ChatGPT Conversations Have Been Indexed by Google

Your product might show up in an AI-generated answer — and then get indexed by Google. This article dives into what that means for brand control and public perception.

Gen Z Social Search Shift: How Brands Must Adapt

If younger customers are searching on TikTok and Instagram instead of Google, your AI strategy needs to stretch beyond product feeds and schema. This article shows how to keep up.

Frequently Asked Questions

What is AI search optimization?

AI search optimization is the process of making sure your products and brand show up in answers from AI tools like ChatGPT, Perplexity, and Google Gemini. It involves things like schema markup, product feeds, natural-language content, and third-party mentions — not just old-school SEO.

Why does AI search matter for ecommerce?

Because more shoppers are using AI to research before they buy. If your products don’t show up in those conversations, you lose visibility — even if your website is technically sound. AI tools recommend based on prompts, context, and brand relevance, not just keywords.

How do I know if AI bots are crawling my site?

You can check your server logs or use your CDN (like Cloudflare) to track requests from bots like GPTBot, PerplexityBot, and Googlebot. This tells you which pages are being indexed — and whether your product pages are being seen.

Do I need to update my product pages?

Yes, especially if they rely on JavaScript to load key content like prices, descriptions, or images. AI tools can’t read content that’s not in the raw HTML. Make sure your product detail pages are readable even with JavaScript turned off.

What is schema markup and why is it important?

Schema markup is structured data that helps AI and search engines understand your product content. It improves visibility in AI-generated answers by clearly explaining what your product is, how much it costs, whether it’s in stock, and more.

Should I submit product feeds to AI platforms?

Absolutely. Submitting your feed to platforms like Google Merchant Center, Perplexity, and (soon) OpenAI helps ensure your products are eligible to show up in AI shopping results. Feeds act as the “source of truth” for product info across many platforms.

How is prompt-based search different from keyword search?

Prompt-based search reflects how people naturally ask questions, like “What’s a good gift for someone who hates clutter?” rather than “minimalist gift ideas.” AI tools look for content that answers full prompts — so your product descriptions and site copy should mirror that language.

What’s the best way to track AI visibility?

Build a library of real prompts your customers might use, and then test whether your brand shows up using tools like Semrush’s AI SEO Toolkit or Peec.AI. Traditional rankings matter less — what matters now is whether you're part of the answer.

How do I get my brand mentioned on third-party platforms?

Encourage reviews on places like Trustpilot, Amazon, and Google. Participate in forums like Reddit and Quora (without being spammy), collaborate with YouTubers or bloggers, and look for opportunities to be included in listicles and product roundups.

Where should I start if all this feels overwhelming?

Start with your messaging. If your product descriptions and feeds don’t clearly reflect what real customers are looking for, even the best schema won’t help.Download the 5-Minute Marketing Fix to craft a one-liner that works across websites, ads, and AI tools alike.