The AI Readiness Report: E-Commerce Edition

When we published our State of AI Readiness 2026 report, the finding that shocked people most wasn't the overall average of 57. It was the e-commerce number: 42 out of 100. That's 15 points below the web-wide average and a full 25 points behind Tech/SaaS.

This matters because e-commerce is arguably the category with the most at stake. When someone asks an AI assistant to recommend a product, the assistant pulls from sites it can access, understand, and trust. At a score of 42, most online stores fail on all three counts.

I've worked with e-commerce brands for over a decade, first in iGaming where product presentation is everything, and now across broader retail. The pattern I see is consistent: e-commerce teams invest heavily in paid acquisition and conversion rate optimization but treat technical SEO as an afterthought. With AI agents becoming a significant discovery channel, that technical debt is about to become very expensive.

AgentReady™ scanned 620 e-commerce sites as part of our 5,000-site study. Here's what we found.

Platform choice has a measurable impact on AI readiness, though it's not the whole story. Shopify stores average 38, WooCommerce sites average 45, BigCommerce averages 47, and custom-built stores average 62.

Shopify's low score might surprise people given its popularity, but the causes are structural. Shopify's default robots.txt blocks several paths that AI crawlers use. Its default theme implementations often lack comprehensive Product schema beyond the basics. And Shopify's liquid templating system makes it harder to implement advanced structured data without apps that add their own performance overhead.

WooCommerce scores 7 points higher primarily because WordPress's plugin ecosystem provides more granular control over robots.txt, schema markup, and meta configuration. Sites running Yoast or RankMath with proper schema configuration consistently score in the 55-65 range.

Custom builds score highest because they're typically built by teams with technical sophistication. They implement schema from scratch, configure AI protocols intentionally, and don't inherit the default limitations of platform themes.

The gap isn't random. Four specific patterns explain almost all of the shortfall.

Root Cause 1: Thin product descriptions. 67% of e-commerce pages we scanned had fewer than 150 words of unique content. Product pages that consist of a title, a price, a one-line description, and bullet points copied from the manufacturer give AI systems almost nothing to work with. The Content Quality subscore for e-commerce averages just 34 out of 100.

Root Cause 2: Missing or minimal Product schema. Only 31% of e-commerce sites implement Product schema with all recommended properties (name, description, price, availability, brand, review, image). Most have either no schema or only the bare minimum auto-generated by their CMS. Check our schema markup guide for the full list of properties that matter.

Root Cause 3: JavaScript-rendered content. 44% of e-commerce sites serve product content primarily through JavaScript, which many AI crawlers either can't or don't render. Server-side rendering or hybrid approaches score significantly better.

Root Cause 4: No AI protocols. Only 5% of e-commerce sites have an llms.txt file, compared to 18% for Tech/SaaS. The concept simply hasn't reached most e-commerce teams yet.

The good news is that e-commerce sites have the most headroom for improvement. Because scores are so low, relatively straightforward fixes can produce dramatic gains.

Fix 1: Implement comprehensive Product schema (+12 points avg). Go beyond the defaults your CMS provides. Add brand, aggregateRating, review, offers with availability and priceValidUntil, and image properties. For product categories, add CollectionPage schema. This single fix moves the Structured Data subscore from 38 to roughly 65.

Fix 2: Unblock AI crawlers and add llms.txt (+10 points avg). Audit your robots.txt for blocks on GPTBot, ClaudeBot, and other AI user agents. Then create an llms.txt file that describes your store, your product categories, and links to your most important pages. Follow our llms.txt guide for the exact format.

Fix 3: Enrich product descriptions to 300+ words (+6 points avg). Every product page should have unique, descriptive content that goes beyond manufacturer specs. Include use cases, comparisons, sizing guidance, and answers to common questions. This lifts both Content Quality and Authority subscores.

Combined, these three fixes are worth approximately +28 points, which would move the average e-commerce site from 42 to 70, crossing into B-grade territory.

Shopify's average of 38 sounds dire, but I want to be clear: the platform isn't fundamentally broken for AI readiness. The low score reflects default configurations, not platform limitations.

Shopify stores that install a schema app (JSON-LD for SEO or Schema Plus are the most common), manually edit their robots.txt.liquid file to allow AI crawlers, and add an llms.txt file through their theme's static assets score between 58 and 72 in our data. That's a 20-34 point improvement over the default.

The issue is that these steps require deliberate action. Unlike platforms with richer plugin ecosystems, Shopify doesn't surface AI readiness as a configurable option. Until Shopify adds native AI protocol support (which I expect by late 2026), store owners need to take manual steps.

For the detailed walkthrough, see our robots.txt guide for AI crawlers which includes Shopify-specific instructions.

The current state of e-commerce AI readiness represents a significant first-mover opportunity. When only 5% of stores have AI protocols and only 31% have proper product schema, the barrier to competitive advantage is low.

But this window won't last. I've seen this pattern before in SEO: early adopters of mobile optimization, page speed improvements, and structured data gained disproportionate advantages that were much harder to replicate once everyone caught up. The same dynamic will play out with AI readiness, likely faster.

Our data shows that e-commerce sites that score above 70 receive 3.4x more AI crawler visits than those below 50. As AI shopping assistants become mainstream, that traffic gap will translate directly into revenue. The stores that act now will have established AI authority by the time their competitors realize they need to start.