Live store detection: proof a store is alive, on every record

What live detection actually means

Most store databases are built by inference: a domain looks like a shop, a directory lists it as a shop, so it goes in the file. Inference decays silently. Shops close, domains lapse, storefronts get replaced by parked pages — and the list keeps selling the corpse as a customer.

Live detection replaces inference with an observation. OpenData's scanner requests the storefront itself and reads what comes back. If the page carries at least one detectable ecommerce technology — a platform signature, a cart or checkout system, a payment integration — the store is verified as alive and selling at the moment of the fetch. If it does not, it is not sold as verified, whatever a directory once said about it.

That single rule is why OpenData's headline number is a verified count, not a scraped count: 1,504,744 verified stores out of 23,984,359 domains scanned (snapshot 15 Nov 2026). The gap between the two numbers is the point — scanned means we swept the domain, verified means we watched it sell. The full counting model is explained in our ecommerce store database guide.

The outcomes that are not 'verified'

A detection pass produces more than one outcome, and honest data work means counting all of them instead of quietly discarding the inconvenient ones:

Each outcome stays in its own bucket. Nothing is silently dropped and nothing is promoted without evidence — a domain only crosses into the verified set when a fetch proves it.

• Verified — the storefront loaded and at least one ecommerce technology was detected.
• Alive, not a store — the site responds, but nothing on it sells anything: blogs, agencies, parked brands, brochure sites.
• Dead — the domain no longer resolves or the server no longer answers.
• Blocked or no response — the store refused or failed the fetch, so its state is unverified rather than assumed.

Every record carries its own evidence

Detection would mean little if the proof stayed in our internal logs. It does not. Each verified record in an OpenData file carries the source URL the data came from and the last-observed date — the day our scanner last saw that store alive with its technology in place.

That changes how you can use the data. Before an outreach campaign, sort or filter by last-observed date and work the freshest segment first. Spot-check any record by opening its source URL yourself: what you see is what the scanner saw. And when a store does close, its next failed check retires it — the record's history is evidence, not decoration. The fields detection and enrichment produce are listed in full on the full column spec page.

Re-checked before every packaging run

One clean pass is not enough, because stores keep dying after they are verified. So detection is not a one-time gate at import: before every refresh run, stores are re-checked, and the ones that no longer respond are dropped from the file before it is rebuilt.

The data is refreshed monthly, and every pull draws on the current snapshot — so the slice you work from keeps being re-proven, not just re-sold. It is also why the free store checker matters: look up any domain and see its current detection state before you spend anything, with 30 full-record unlocks a month on a free account.

The checker shows the same classification the packaging pipeline uses — verified, alive-not-store, dead or unverified — not a softened marketing subset. If a domain is in the verified set, its record shows the source URL and the day it was last observed; if it is not, the checker says plainly what it actually is. One index, one standard, whether you pay or not.

What this means for a buyer

Practically, live detection buys you three things. First, deliverability of effort: your outreach, research or integration work targets stores that existed when last observed, not when some crawler guessed. Second, auditability: source URL plus last-observed date on every record means any claim in the file can be checked in seconds, record by record. Third, honest coverage: the 269,688 verified Shopify stores in the database are stores a scanner watched operating on Shopify — the same standard applies to every platform slice.

Detection is the first half of the record. The second half — platform, apps, payment gateways, contacts — is layered on top by data enrichment, and the two together are what make a raw domain into a usable store record.

Questions, answered straight

Next steps