Shopify
6.4%
20 of 320 stores had a tag loading and staying silent on at least one page type whose control signal was satisfied.
How many live stores have broken tracking right now
500 stores scanned in September 2026, sampled at random. Method and exclusions stated in full.
WooCommerce
11.3%
20 of 180 stores. Plugin churn is the likeliest explanation and we did not test for it.
A store with no analytics, and a store we could not observe, is not in these rates. A consent banner that never resolved is unobservable, not broken. Including the unobservable would make the headline look healthier than the web.
By vendor
Presence is not the same as firing on every page type we could load.
| Vendor | Stores present | Firing everywhere | Uneven | Loading, silent |
|---|---|---|---|---|
| Google Analytics 4 | 418 | 371 | 28 | 19 |
| Meta pixel | 296 | 248 | 33 | 15 |
| Google Ads | 201 | 182 | 12 | 7 |
| Klaviyo | 134 | 96 | 34 | 4 |
| TikTok | 88 | 71 | 11 | 6 |
The CSV for this run is not published yet. The counts above are the ones the homepage cites, and they are the ones a later file will have to match. Until then,read how the scanner works, including what it cannot see.
How the sample was drawn
Random hosts from public Shopify and WooCommerce storefronts. A store enters the sample once; it is not weighted by traffic. English-language stores are over-represented. A host that blocks the scanner is recorded as unobservable rather than replaced, so the frame is not silently healthier than the web.