Episode 3: If AI Can’t Read Your Product Page, It Doesn’t Exist

‍

If you haven't heard the term Generative Engine Optimization (GEO) yet, it's time to get familiar. Think of it as SEO for 2026 — the practice of making your product pages and data visible to AI-powered search tools like ChatGPT, Perplexity, Google AI Overviews, and others.

The numbers tell the story: Shopify has reported up to 15x more sales coming from AI agents, and around 70% of customers are now using AI as part of their shopping experience. This isn't a niche shift — it's mainstream.

After running AI readiness audits across businesses in automation, medical supplies, education, and building supplies, we keep seeing the same issues come up. The good news? They're fixable. Here are the three most impactful things you can do right now.

1. Make Sure Your Product Data Is Actually in the HTML

This one surprises a lot of teams. Your product data might live in a CMS or PIM — but if it's not rendered in the actual HTML of your product page, AI crawlers can't reliably read it.

The key elements that need to be visible in plain HTML include:

• Descriptive product titles
• Written product descriptions
• Technical specifications
• Product variants

A common culprit is third-party apps or widgets that load content dynamically via JavaScript. Some AI crawlers can index JavaScript content, but they de-prioritise it because it loads slower and can be inconsistent.

A quick way to check: do a View Page Source on your product pages. If you can't read your product info there, neither can an AI crawler — and that means your data is effectively invisible.

For Shopify sellers in particular, avoid loading product descriptions through dynamic third-party apps where possible. If your setup does rely on JavaScript rendering, talk to your dev team about server-side rendering as a fix.

The bottom line: if AI can't read it without JavaScript, it might as well not exist.

2. Add and Validate Product Schema

Product schema is a structured way of marking up your page so that machines — and AI systems — can clearly understand what your product is and what it offers.

While schema doesn't feed directly into LLMs the way Google's traditional search index does, it strengthens the signals that AI systems rely on. It feeds into Google's index, Merchant Center Shopping feeds, and Knowledge Panels — all of which are key sources that tools like Gemini and Perplexity cross-reference when generating answers.

At a minimum, your product schema should include:

• Product name / title
• Brand
• Description
• Images
• Product codes and MPNs
• For technical products: regulatory compliance info, materials, and spec data

You can find full implementation guidance in Google's Search Central documentation under product schema.

One important note: many e-commerce platforms like Shopify and WooCommerce generate schema automatically — but that doesn't mean it's working correctly. Always validate that your schema is actually being applied and rendering properly across your product pages. Don't assume plugins are doing their job without checking.

3. Write Content That Answers Real Questions

The way people search has changed. Instead of searching for "red trainers," customers are now asking things like "What's the most comfortable lightweight shoe for standing all day?" That shift in search behaviour means your product content needs to shift too.

The highest-scoring product pages we've audited share one thing in common: they have a narrative around the product, not just a data sheet. That means content like:

• Use case paragraphs — who is this product for, and what problem does it solve?
• Comparison language — how does this differ from alternatives, and when would you choose it?
• FAQs — answer the questions your customers are actually asking
• Descriptive image alt text — images are processed as part of AI knowledge bases, so make alt text count

One shift worth flagging for brands: highly stylised, emotive copy ("walking on clouds") doesn't translate well for AI. Direct, descriptive language ("lightweight, cushioned shoe for all-day comfort") gives AI the context it needs to surface your product in relevant searches. You don't have to strip out your brand voice entirely — but make sure the substance is there underneath it.

And creating this kind of content at scale is increasingly achievable. With today's LLM tools, you can generate long-form product content quickly — as long as you have solid, structured foundational data to start from. Which, of course, brings everything full circle.

Quick Recap

‍Static HTML: If AI can't read it without JavaScript, it doesn't exist. Render your product data in the page source.

‍Product Schema: Structured, machine-readable, validated, and consistent. Gives AI the context to place your product in search.Question-Answering ContentUse cases, comparisons, FAQs, alt text. Every paragraph should be able to stand alone as an answer.

If you'd like us to take a look at your product pages and run an AI readiness audit, feel free to reach out. You can also try it yourself — run some prompts against your pages and see what an AI can find.

‍