Download websites that use Insider

About Insider

By the time a personalised page reaches a visitor, parts of it have been chosen for that visitor. Recommendations, banners, badges and sometimes the message itself are assembled at load time from a behavioural profile. Insider is one of the systems doing that assembly, which makes the tag an active participant in what the page says rather than a passive beacon watching what the visitor does.

Buying a platform of this kind is a commitment rather than a purchase. Somebody has to feed it a product catalogue, define segments, run tests and keep campaigns alive, and the software is worth nothing without that continuing work. Detection therefore identifies a company with a marketing team, a budget line for retention rather than acquisition alone, and enough traffic for experimentation to mean anything. Very small storefronts do not appear here, not because they are excluded but because the arithmetic never works for them.

Insider has sold the platform since 2012 under a Proprietary model, mainly to retail and ecommerce brands, and the Customer data platform classification places it beside tools that overlap heavily in function. Overlap is the useful part. A site running this platform has usually already decided against building personalisation in house and against at least one competing suite, so the stack around the tag records a sequence of decisions rather than a single one.

As of September 2026 we detect Insider on 3,919 sites. The tag is JavaScript loaded in the browser, which is what makes it visible to a crawler, with two familiar qualifications attached. Deployment through a tag manager or behind a consent wall can suppress it during a single unauthenticated fetch, and larger customers frequently serve the script from a subdomain of their own, which defeats naive host matching. Where possible we match on script behaviour and path patterns as well as on host, but a first party served tag will always be harder to see than a third party one, and enterprise accounts are the ones most likely to be served that way.

The full list matters because every interesting question in this category is comparative. A competing vendor wants the accounts a rival already holds, filtered by country and platform. A partner or agency wants brands running the platform without the complementary tooling that usually accompanies it, a gap visible only by joining across technologies. An investor wants the direction the installed base is moving in. None of that survives being asked one domain at a time. A downloadable dataset delivers the population and the surrounding stack together, which is the only form in which those questions have answers at all.