Download websites that use Sift

About Sift

Fraud scoring is back office work. A risk engine weighs the card, the device, the shipping address and the age of an account, returns a verdict, and the buyer never learns it happened. Sift sells that kind of engine, which makes an appearance in a website technology survey slightly counterintuitive. What a crawler can see is only the front edge of the product, a collection script that gathers device and behaviour signals from the browser and hands them to the vendor for scoring. Everything downstream of that script, the models, the rules, the manual review queues, leaves no trace in the markup at all.

As of October 2026 we record the collector on 2 sites. A figure that small should be read as a floor rather than a census, for two reasons that both push in the same direction. Fraud tags are frequently loaded only on the pages where money or identity is at stake, meaning checkout, login and registration, so a crawl that reads a homepage will simply miss them. Larger deployments also tend to proxy collection through a first-party host name or fire the tag from a container, which hides the vendor behind the customer's own domain.

Presence is worth something even with the undercount. A domain carrying this Analytics script belongs to a business where fraud is expensive enough to justify a dedicated platform instead of the screening built into a payment processor. Marketplaces, subscription products, digital goods and anything holding stored value tend to cross that threshold early. The platform is a Proprietary offering from Sift and has been sold since 2011, long enough that adoption reflects a considered purchase rather than a trial.

Two audiences use a list like this. Competing risk vendors want the installed base as a prospect map, because displacement in fraud tooling is slow and knowing who already made the decision matters more than raw traffic figures. Integration partners, payment gateways, identity providers and commerce platforms want the reverse view, the set of merchants whose stack their own connector should support.

A single domain lookup answers whether one merchant uses the platform. The downloadable list answers the questions that need the whole population at once: what those merchants sell, which commerce systems they run on, which country domains they trade under, and how the set changes between releases. Cross-referencing is the point, and cross-referencing needs the file rather than a query.