Agentic Commerce Readiness by Platform (2026)
TL;DR: We ran our free agentic-commerce check against 23 real storefronts — four each on Shopify, WooCommerce, BigCommerce, Wix and Squarespace, three on Webflow. Every single one landed at level 0: invisible. Not one of the 23 could be found, understood and bought from by a shopping agent acting on a customer's behalf. The more useful finding is the spread: all four Shopify stores scored identically, and most platforms clustered on one or two values. On this, the platform sets your score and your merchandising barely moves it.
Key Facts:
- 23 storefronts graded on 2026-09-14. Every one at level 0 of 4 — invisible to a buying agent.
- All four Shopify stores returned the same score. Wix, WooCommerce, Squarespace and Webflow each returned only two distinct scores across their stores.
- BigCommerce was the only platform with real variation, and held the single highest score of all 23.
- Every store was confirmed to be on the platform we credit it to before being graded — several widely-cited "built on X" case studies turned out to describe stores that have since moved to a different platform.
- The check is free, reads only public pages, and asks no AI engine anything.
What "readiness" means here, in one paragraph
There is a difference between a customer finding your site and a customer's agent completing a purchase on it. We grade that second thing on five levels:
| Level | Name | What it means |
|---|---|---|
| 0 | Invisible | An agent cannot reliably tell what you sell |
| 1 | Listed | Your products are discoverable, but not machine-readable |
| 2 | Discoverable | An agent can read your catalogue |
| 3 | Buyable | An agent can complete a purchase |
| 4 | Operated | An agent can transact, track and resolve without a human |
Twenty-three stores. All of them at zero.
The results
| Platform | Stores graded | Score range | Median | Distinct scores |
|---|---|---|---|---|
| BigCommerce | 4 | 4–28 | 12 | 3 of 4 |
| Shopify | 4 | 12 | 12 | 1 of 4 |
| Wix | 4 | 8–12 | 8 | 2 of 4 |
| WooCommerce | 4 | 4–8 | 8 | 2 of 4 |
| Webflow | 3 | 0–4 | 4 | 2 of 3 |
| Squarespace | 4 | 0–4 | 0 | 2 of 4 |
Every store: level 0.
The number that matters is the last column
Four Shopify stores. Different industries, different catalogue sizes, different agencies, wildly different budgets. The same score.
That is not four merchants who happened to make the same choices. That is a platform ceiling. Whatever those four teams did with their themes, their copy, their structured data and their apps, none of it moved the thing an agent needs. The same clustering shows up on Wix, on WooCommerce, on Squarespace and on Webflow — one or two values across every store we graded.
If you are on one of those platforms, this is the useful and slightly uncomfortable news: your agentic readiness is currently mostly not your doing. You inherited it.
The exception, and what it suggests
BigCommerce was the only platform where the stores genuinely disagreed with each other — a range from 4 to 28, and the highest single score in the whole sample. Three distinct values out of four stores.
A spread means merchant decisions are reaching the surface an agent reads. A cluster means they are not. Of the six, BigCommerce was the only platform where what the merchant did appeared to matter.
We would not build a purchasing decision on four stores, and neither should you. But it is the one result in this sample that points at a difference between platforms rather than a difference between merchants.
Why every score is low, even the high one
The top score in the sample was 28 out of 100 — and that store is still at level 0.
Shopping agents are new. The conventions that let one read a catalogue, confirm a price, check a returns policy and complete a purchase without a human are arriving now, and essentially nothing that was built before them speaks that language. A site can be beautiful, fast, well-ranked in Google and commercially successful while being, to an agent, a wall.
That is the whole opportunity. Level 0 is not a failing grade handed out to laggards. Right now it is the market.
What to do about it
If you are choosing a platform: don't choose on this. Every one of the six is at level 0 today. Choose on everything you would normally choose on, and treat agentic readiness as work you will do on top.
If you are already on one: the gap is not your theme and it is not your copy. It is that nothing on your site tells a machine what you sell, what it costs, and what happens if it is sent back — in a form the machine can act on rather than render. That work is largely independent of which of these six you are on.
If you sell to merchants: the clustering is your opening. When every store on a platform scores the same, the first one that doesn't is visible to every agent looking.
How we did this
We graded storefronts we confirmed were on each platform — not stores an article claimed were on it. That distinction cost us four of our first six picks: widely-repeated case studies named stores that have since moved to Shopify. Each store was then graded with the same free check anyone can run against their own domain, reading only public pages.
We deliberately grade several stores per platform. Our first pass graded one each, got six results at level 0, and could not tell us whether that was six platforms or six merchants. It took four per platform to see the clustering — which turned out to be the finding.
Numbers are from 2026-09-14 and will move. The dated record behind this table is kept alongside the post.
Run the same check on your own domain · Method and its limits · What an agent needs to see
Does AI name your site? Free 60-second check.
See who ChatGPT, Perplexity, and Google’s AI name for your buyers’ questions.
Written with AI assistance and reviewed by LoudPixel for accuracy.
Check your AI search visibility.
See which AI engines name your site — in about 60 seconds. Free, no signup.