Hotjar in the Consent Gap Index
Hotjar is a behaviour analytics product. It records how visitors move through a page — where the cursor goes, how far they scroll, which elements they click — and plays that back to the site owner as heatmaps and session replays. You would meet the name in a page's script list, in a cookie notice, or in a content blocker's report. This index found it on 178 sites, 3.4 percent of the 5,294 measured. The sites that run it carry the heaviest median third-party load of the 13 technologies indexed here.
Ten companies is the highest median in the set
The median site running Hotjar contacts 10 distinct companies, of which 3 are classified as trackers. Both of those are the highest medians among the 13 technologies in this index — rank 1 of 13 on each. Across 48 requests to third-party hosts, which ranks 5th of the 13, that is a page with a genuinely crowded edge. The interpretation is not that Hotjar pulls the other nine companies in. It is that the decision to install session-replay software tends to be made by teams that have already installed a tag manager, an ad pixel, a support widget and an A/B testing tool. Hotjar is a marker for that kind of site rather than a cause of it.
Heavy, but slow off the line
The pattern that makes this group unusual is that the heaviest stacks in the technology set are also among the slowest to start. The median Hotjar site made its first third-party request at 766 milliseconds — the 3rd-latest median of the 13 technologies here. The fastest first contact recorded on any Hotjar site was 75 milliseconds. Early contact happened on 156 of the 178 sites, 87.6 percent, which ranks 11th of 13: two technologies in this index have a lower share, ten have a higher one. So the group with the most companies per page is also the group where those companies are relatively less likely to have been reached in the opening window.
Seventy-seven of 178 sites carried a banner
A consent tool was present on 77 of the 178 Hotjar sites. Six were seen alongside them: OneTrust on 30, an IAB TCF consent tool on 29, Usercentrics on 5, Osano on 4, CookieYes on 3 and Cookiebot on 3. On the median Hotjar site, 6 companies were contacted before an answer could be registered and 2 more after the consent window closed. That second figure is worth reading on its own: a Hotjar page tends to keep adding parties once the opening burst is over, which is the expected behaviour of anything that starts reporting only after a page has settled and a visitor has begun to move around it.
The spread across twelve named sites
tandfonline.com sits at the extreme: 10 companies contacted, all 10 of them early, first contact at 869 ms, behind an IAB TCF banner. cambridge.org shows 7 of 7 early at 560 ms and pexels.com 6 of 6 at 293 ms, both behind OneTrust. bugsnag.com shows 5 of 5 at 710 ms with no banner recorded, and stanford.edu 3 of 3 at 1,592 ms. Then there is the other tail. dnsmadeeasy.com contacted 7 companies with none of them early and a first contact at 3,153 ms; mercadolivre.com.br contacted 3, none early, at 2,261 ms. fwmrm.net is the heaviest single site in this sample at 14 companies, of which 5 were early.
Middle cases
Most sites are neither extreme. discogs.com contacted 6 companies with 4 early at 1,189 ms; dreamhost.com 5 with 2 early at 866 ms; mediafire.com 6 with 3 early at 1,173 ms; shopee.co.id 4 with 3 early at 1,292 ms. Half or so of the parties arrive in the opening window and the rest follow. When people picture what a tracked page does, they usually picture the tandfonline.com case — everything at once, up front. The median is closer to these: a first wave, then a long tail that continues while the page is being read.
Reading a session recorder in context
Session replay is the one category on a normal page where the collected material is closest to a recording of a person rather than a count of them, which is why it draws more scrutiny than its site count would suggest. What this index can say about it is narrow and specific: 178 sites, the highest median company and tracker counts of the 13 technologies measured, a late median first contact, and 77 sites with a banner.
| Site | Companies | Before consent | First contact |
|---|---|---|---|
| dnsmadeeasy.com | 7 | 0 | 3153 ms |
| stanford.edu | 3 | 3 | 1592 ms |
| tandfonline.com | 10 | 10 | 869 ms |
| dreamhost.com | 5 | 2 | 866 ms |
| mercadolivre.com.br | 3 | 0 | 2261 ms |
| cambridge.org | 7 | 7 | 560 ms |
| mediafire.com | 6 | 3 | 1173 ms |
| shopee.co.id | 4 | 3 | 1292 ms |
| bugsnag.com | 5 | 5 | 710 ms |
| pexels.com | 6 | 6 | 293 ms |
| discogs.com | 6 | 4 | 1189 ms |
| fwmrm.net | 14 | 5 | 1088 ms |
| Tool | Sites |
|---|---|
| OneTrust | 30 |
| IAB TCF consent tool | 29 |
| Usercentrics | 5 |
| Osano | 4 |
| CookieYes | 3 |
| Cookiebot | 3 |
The six most common are listed. The remaining 3 of 77 sites with a banner carry something outside that six. One tool is recorded per site, so these counts never overlap.
Questions
What does Hotjar collect?
It is a behaviour analytics tool: heatmaps, scroll depth, click positions and session replays of how a visitor moved through a page. This index measures which companies a page contacts and when, not the contents of what any one of them receives.
Why do Hotjar sites contact so many companies?
Selection, most likely. A median of 10 companies is the highest of the 13 technologies indexed here, but installing session replay is a decision made by teams that generally already run a tag manager, ad pixels and testing tools. The tool marks a type of site.
Does the median Hotjar site show a consent banner?
No. 77 of 178 carried one, so slightly fewer than half. OneTrust and an IAB TCF tool were the two most often seen alongside them, on 30 and 29 sites respectively.
How many sites are in the full index?
5,294 sites were measured for this snapshot. Hotjar was found on 178 of them, ranking 7th of the 13 technologies here by site count.