We scored 245 public Shopify storefronts on how much of them an AI assistant can actually read. The median came out at 46 out of 100, and not one store reached the top band.
- Median Shopify score: 46 out of 100. Two thirds of stores sit below half marks.
- Shopify stores score lower than the general web, which scored 47.2 across 613 sites.
- No website in our benchmark had reached the Leading band when we published this on August 22, 2026.
What did we measure?
Every site is fetched the way an assistant fetches it, then scored from 0 to 100 on how much of it can be read, understood and quoted. Same engine, same methodology version, no platform special casing, no manual adjustment.
The numbers below were read from the live benchmark on the day this was published rather than from a spreadsheet made months ago. They move as the cohort grows, which is the point of publishing the method alongside the result.
How did Shopify stores score?
Poorly, and consistently so. The distribution is narrow and it sits low.
| Measure | Shopify storefronts | All websites |
|---|---|---|
| Sites scored | 245 | 613 |
| Median score | 46.0 | 47.2 |
| Scoring below 50 | 67% | 61% |
| Highest score reached | 69.2 | 71.5 |
| Top decile starts at | 55.9 | 58.8 |
Do Shopify stores do worse than the rest of the web?
On this measure, yes. A lower median, a larger share below half marks, a lower ceiling and a lower top decile. The gap is not enormous, but it runs the same direction on every measure, which is harder to dismiss than a single number.
That is a surprising result for a platform that handles the technical basics well. Shopify gives every store a working sitemap, canonical tags and a fast responsive theme, and none of that is where stores are losing. They lose on the layer above: structured data that matches the page, product attributes stated as data rather than implied, and copy an assistant can lift as an answer.
What do the bands look like?
The shape of the distribution says more than the median does.
Why is the top band empty?
Because reaching it requires doing every layer well at once, and almost nobody is doing the last layer at all. A store can have flawless technical foundations and still be unquotable, because being fetchable and being useful to an assistant are different achievements.
The empty band is the most useful finding here. It means the ceiling in this category has not been set by anybody yet, and the work required to be the best readable store in a niche is currently small.
What should a store owner take from this?
That the bar is low, which is good news if you move now. A store scoring 60 is already in the top few percent of its platform, and 60 does not require anything exotic.
Run the same scan on your own store and compare against these numbers directly. The manual version of these checks takes twenty minutes if you would rather do it by hand first.
Method, so you can check us
Cohort: every distinct public domain scored at methodology version aivs-1.0-aeo20260620-full-evid20260621. Shopify cohort n=245, full cohort n=613, the latter including ecommerce and non ecommerce sites scored under the same rubric.
When we score the same unchanged site twice, the two scores usually land within about 2 points of each other. We publish our own score under the same engine rather than exempting ourselves, which is at our own score.
Common questions
Is 46 out of 100 bad?
It is below the level where assistants reliably describe a store accurately. It is also completely normal, which is the point. Two thirds of Shopify stores are below 50.
Why did no store reach the Leading band?
Reaching it needs every layer done well simultaneously. Most stores do the technical layer, which the platform largely handles, and stop before the layer that decides whether they can be quoted.
Can I see my own store against these numbers?
Yes. The free scan uses the same engine and the same methodology version, so your score is directly comparable to the medians above.
Will these numbers change?
They already are. The cohort grows as more sites are scored, and the figures here were read live on the publication date rather than frozen months earlier.
Related reading
Run a free scan and get a 0 to 100 score with a fix list in priority order. Run the free scan.
Your buyers are already asking AI. This is how you make your website readable to the assistants they ask.
Everything in this article is measurable on a live storefront. The Full Assessment and Roadmap reads your website the way AI crawlers receive it and hands you every fix in plain English, in the order we would make them.