Companies Using Crazy Egg

This profile was rebuilt from scratch in September 2026. The first thing to explain is why its headline number moved, because if you saw this page earlier in the year it said something very different.

1. What changed on this page

Crazy Egg was previously tracked here under a legacy technology id that our current detection cluster does not contain at all. The profile has been re-pointed to the live id, re-measured and re-filed: the count moved from 28,857 to 122,407 domains, and the category moved from Web Analytics to Heatmaps & Session Recording.

The 4.2x difference is not adoption. The two figures come from different detection systems measuring different things, with no growth event between them, and nobody should build an argument on the gap. The category change is the more meaningful of the two corrections. Filed under Web Analytics, Crazy Egg was ranked against several hundred general analytics tools and looked like a rounding error at a fraction of a percent. Filed where the vendor's detection rule actually sits, it is third of 34 with a 7.2% share. Same product, same install base, different denominator - which is a useful reminder that a market share percentage is a statement about the category boundary at least as much as about the product.

2. How Crazy Egg is detected

Two signals identify it. The first is the JavaScript global CE2, which the Crazy Egg snippet defines. The second is a script source matching script\.crazyegg\.com/pages/scripts/\d+/\d+\.js. The second signal is the stronger of the pair: it is a vendor-owned host with a structured path, so a false match is hard to produce by accident. Both are page-surface signals, which sets the limits of everything below. A detection means the snippet was served on the page we fetched. It does not tell us whether the account behind it is paid, active, or being looked at by anyone, and no figure on this page should be read as a customer list.

3. The install base is small businesses, and old ones

44.25% of enriched customers have 1-10 employees and 74.34% have fewer than 50. Only 0.75% have 10,001 or more. Geographically it is an English-language product: 20,681 of 32,366 enriched companies are in the United States, and the UK, Canada and Australia take the next three places.

The founding-decade split is the part worth pausing on. The 2010s dominate at 32.44% and the 2000s add 20.73%, but the 2020s contribute only 7.65% - while businesses founded before 1960 account for 10.45%, more than the newest cohort. Crazy Egg is not what a company founded last year reaches for; it is what an established small business installs once it has enough traffic to argue about a page. That is a retention story and an acquisition problem at the same time.

4. Almost every Crazy Egg site runs Google Tag Manager

Google Tag Manager is present on 98.36% of Crazy Egg's domains and Google Analytics on 97.77%. Those numbers are far outside the normal band for either pairing, which usually lands somewhere between 45% and 60%, and a co-occurrence that extreme is worth interrogating before it is narrated, because the usual cause is a shared host, a template or a single agency's client fleet rather than anything about the product.

It is not that here. A random sample of detected domains was checked for a common platform: it returned independent small American businesses in unrelated trades - a church, car dealerships, a container supplier, estate agents, a staffing firm, a trucking company, retailers - on unrelated hosts, with no shared subdomain pattern between them. The base is genuinely independent.

The real explanation is the install path. Crazy Egg is most commonly deployed through Google Tag Manager rather than hard-coded into the page. The vendor documents a one-click GTM integration, it is the route that does not require a developer, and it is therefore the route a marketing team takes. A site running Crazy Egg nearly always runs GTM because GTM is usually how Crazy Egg got there; a site running GTM nearly always has Google Analytics in the same container. This says something about who buys the product - a marketer with container access, not an engineering team - and nothing at all about a shared platform.

5. Hotjar is a neighbour more often than a replacement

Hotjar is Crazy Egg's closest competitor by product and by size, at 440,057 domains to Crazy Egg's 122,407. It is also detected on 64.31% of Crazy Egg's own domains. Nearly two thirds of the sites running Crazy Egg are running its nearest rival alongside it.

This is why the profile carries no migration table. A switching corridor is only publishable when the data shows one product displacing another, and displacement cannot be inferred between two products that so plainly coexist - a count of sites that "moved" between them would mostly be counting sites that kept both, or sites where one tag was added for a single experiment and never removed. Across this refresh programme the only corridors to survive testing have been ones where two products cannot physically coexist, such as competing SEO plugins on a single WordPress install. Behavioural analytics is the opposite shape: the tools are cheap, narrow, and each answers a slightly different question, so stacking them is rational.

The competitive pressure in this category is real, but it comes from price and defaults rather than from head-to-head replacement. Microsoft Clarity holds 54.3% of the category on 923,836 domains and costs nothing.

6. Churn

10,258 domains have dropped Crazy Egg since coverage of it began in March 2026, roughly 8% of the current base. For a JavaScript tag that is a low figure, and it is worth being explicit about what it does and does not mean. A tag disappears when somebody removes it, when a site is rebuilt, when a container is cleaned out, or when the site itself goes away - a drop is not a cancellation, and we do not claim to know where any of those 10,258 went. The useful comparison is against tools whose detection depends on something fragile, such as a consent-gated tag or a rule that matches a version string; those can churn several times harder without a single customer changing their mind. Crazy Egg's signal is stable, so its churn number is closer to a real behaviour than an instrument artefact.

7. Reading the company table correctly

The named companies on this page - AstraZeneca, RBC, Philips, Electronic Arts, the World Bank, Novartis, a long list of universities - are all in the 10,001+ employee bracket, which is 0.75% of this install base. They are the exception the enrichment surfaces, not the customer. Look at the hosts rather than the names and a second pattern appears: careers.astrazeneca.com, jobs.rbc.com, help.ea.com, blogs.worldbank.org, news.umich.edu. Roughly half of the large-organisation detections are on a subdomain - a careers portal, a newsroom, a help centre, a campus or events property - not the corporate homepage. Heatmap tooling enters a large organisation the same way it enters a small one: one team, one property, one card. It does not imply an enterprise agreement, and it should not be quoted as one.

8. Key takeaways

1. The headline moved for measurement reasons. 28,857 and 122,407 come from different detection systems and different categories. Neither is evidence about adoption.

2. Third of 34, in a two-giant category. 7.2% behind Clarity at 54.3% and Hotjar at 25.87%, and 3.6x the fourth-placed tool.

3. The GTM number is an install-path artefact, not a platform. 98.36% GTM and 97.77% Google Analytics describe how the snippet gets on the page; the underlying domains are independent businesses.

4. Coexistence, not competition, with Hotjar. 64.31% overlap is why no migration corridor is published here.

5. Small, established, English-speaking. 74.34% under 50 employees, 20,681 of 32,366 enriched companies in the US, and a founding profile that skews to the 2010s and earlier rather than to new companies.

9. Sales applications

Who to approach. The 98.36% Google Tag Manager overlap is the most actionable line in this dataset, because it identifies the buyer rather than the company. A Crazy Egg detection effectively marks a marketing team with container access and a habit of adding measurement tools without an engineering ticket. Layer HubSpot (66.21%) on top and you have a conversion-focused marketing stack with a demonstrated willingness to buy small tools quickly.

Where the volume is. Filter on the industry mix rather than the logo wall: real estate (5.51%), retail (4.53%) and construction (3.71%) are the three largest verticals, all of them populated by businesses with fewer than 50 staff and a single page that matters commercially. The recognisable enterprise names are a thin tail and a slow sale.

What not to claim. Do not pitch against a migration story on this page - there isn't one, and two thirds of these sites are running Hotjar too. Do not read a subdomain detection at a large organisation as an enterprise relationship. And do not quote the June-to-September rise as growth: our corpus went from roughly 201 million to 379 million active domains over that window, so the curve is mostly our own coverage catching up.