By Nick Quirk
October 6, 2026
9 min read
A heatmap is a colour-coded picture of where visitors click, move their cursor and stop scrolling on a single page, built from hundreds or thousands of real sessions. Both Hotjar and Crazy Egg will produce one within about ten minutes of pasting a script onto your site, and the free or entry tiers are enough to diagnose most landing pages. The hard part is not installing the tool, it is reading the map without talking yourself into a conclusion the data never supported.
Most comparison articles stop at a feature table. This one covers what the colours actually mean, how much traffic you need before the picture is trustworthy, and how to convert a visual into a test you can measure.
The Four Map Types and What Each One Measures
Heatmaps are not one report. Both platforms generate several overlays from the same tracking script, and each answers a different question about user behavior.
- Click maps show where people tap or click, including clicks on things that are not links. Those “dead clicks” are often the most useful signal on the whole page.
- Scroll maps show the percentage of visitors still present at each pixel depth. A sharp colour break reveals a false bottom where the layout looks finished but is not.
- Move maps track cursor position on desktop. Treat them as weak evidence: cursor and eye position correlate loosely, not reliably.
- Confetti or segmented views plot every individual click as a dot, coloured by source, device, campaign or new versus returning visitor. Crazy Egg’s confetti report is the clearest implementation of this idea.
Crazy Egg packages these into what it calls Snapshots, which run for a set period or pageview count and then freeze. Hotjar keeps heatmaps continuously updating and pairs them with session recordings and on-page surveys inside the same dashboard.
How Much Traffic a Heatmap Needs Before You Trust It
A heatmap built on 80 sessions is decoration. Aim for 2,000 to 3,000 pageviews per device type before you treat a pattern as real, and never blend mobile and desktop into one map, because the layouts differ and the averaged result describes nobody.
If a page only gets 300 visits a month, let the snapshot run for six to eight weeks rather than drawing conclusions in week one. Low-traffic pages are usually better served by watching fifteen to twenty session recordings end to end, where a single confused visitor tells you more than a pale cluster of dots.
One more sampling trap: both tools capture a percentage of sessions on lower plans. Check what your plan actually records before you compare two pages that were sampled at different rates.
Reading a Scroll Map Without Fooling Yourself
The number that matters on a scroll map is the average fold, the depth where roughly half your audience has dropped off. On most content pages that lands somewhere between 50 and 65 percent of page height, though long-form posts with strong intros hold readers deeper.
Falling attention down the page is normal, not a failure. Nielsen Norman Group’s long-running eye-tracking work found that people spend the majority of viewing time above the fold and that attention decays steadily as they scroll, which is why placement of your offer matters as much as its wording. Their research on scrolling and attention is worth reading before you redesign anything based on a scroll map.
What should worry you is a cliff: 70 percent of visitors present at one point, 22 percent forty pixels later. That almost always means a horizontal rule, a full-width image or a section of whitespace is reading as the end of the page.
Five Patterns That Justify a Change
Not every hotspot deserves a redesign. These five are the ones that reliably pay for the subscription.
- Dead clicks on non-interactive elements. If 4 percent of visitors click a product image that does not open, add a lightbox or a gallery. People are telling you what they expected.
- Rage clicks. Three or more rapid clicks in the same spot usually means a slow or broken element. Hotjar flags these automatically; in Crazy Egg you spot them in recordings.
- A primary CTA that is colder than the navigation. When your main button attracts fewer clicks than the header menu, the page is leaking people sideways before they ever consider the offer.
- Heavy clicking on a link that goes somewhere low-value. A footer link pulling 6 percent of clicks is a demand signal for content you have buried.
- A form field where cursor movement stalls. Pair the map with a recording and you will usually find a validation message nobody can see on mobile.
Each of these maps to a specific fix, which is the difference between behavioural analytics and a pretty screenshot. If you are working through a wider optimisation programme, these findings slot neatly into the tactics that lift landing page conversion rates.
Hotjar or Crazy Egg: Where Each One Earns Its Place
The two products overlap heavily in 2026, so the choice usually comes down to how you work rather than which feature list is longer.
Hotjar suits teams who want qualitative context attached to the map. Session recordings, on-site surveys, feedback widgets and funnel reporting live in one place, and the free tier captures a limited number of daily sessions indefinitely, which is genuinely useful for a small site. Paid plans typically start near $39 per month and scale with session volume.
Crazy Egg suits teams who want to test as well as observe. Its Snapshots model, confetti segmentation, scroll map exports and built-in A/B testing make it a tighter loop for ecommerce and lead-gen pages, with entry pricing usually in the $29 to $49 per month range on annual billing. Crazy Egg reports being used by well over 449,000 websites, and its surveys and add-on tools cover most of what Hotjar offers on the feedback side.
Mouseflow sits as a third option with friction scoring and form analytics, and Lucky Orange bundles live chat, which matters if support conversations and behaviour data belong together for you. Check current pricing pages directly, since all four vendors have changed their tiers more than once in the last two years.
Segment Before You Interpret
An unsegmented heatmap averages every visitor into a single blurry person. Split the same page by paid versus organic traffic and you will frequently see two different behaviours: ad traffic clicking price and proof elements, organic traffic reading further down and clicking internal links.
Useful segments to build first:
- Mobile versus desktop, always separated
- New versus returning visitors
- Converters versus non-converters, where the plan supports it
- Traffic from a single campaign or UTM
That last one connects behaviour data back to acquisition. If your paid segment behaves nothing like the rest, the fix may be upstream in who you are reaching, which is where refining search campaigns with audience targeting does more good than another button colour test.
Turning a Map Into a Hypothesis You Can Measure
Write the finding as a sentence with a number in it before you touch the page. “Only 31 percent of mobile visitors reach the pricing table, which sits at 78 percent page depth” is actionable. “The page feels too long” is not.
Then state the change and the expected effect: move the pricing table above the testimonial block, expect scroll-to-pricing to rise from 31 percent toward 55 percent and form starts to follow. Run the variant for at least two full business cycles, usually two to four weeks, so weekday and weekend behaviour both get represented.
Keep the before-and-after snapshots side by side in your reporting. Pulling those numbers into a shared dashboard alongside spend and conversions, something Looker Studio handles well for custom reporting, stops heatmap work from living in a silo nobody else reviews.
Privacy, Masking and Data Handling
Session recording tools capture what happens inside a browser, so treat the configuration as a compliance task, not a setting. Both platforms suppress keystrokes in form fields by default, but you should verify masking on checkout, login and any field holding personal or payment data before you record a single session.
Under UK and EU rules, analytics scripts that are not strictly necessary generally require consent, and the ICO’s guidance on cookies and similar technologies sets out the expectation clearly. Wire the heatmap script to your consent banner so it only fires after permission, and note that doing so will reduce your captured sample, which brings you back to the traffic thresholds above.
Frequently Asked Questions
Is Lucky Orange a Better Tool Than Hotjar?
Lucky Orange is usually cheaper at the entry level, with plans commonly starting around $39 per month, and it includes live chat and form analytics that Hotjar does not bundle. Hotjar tends to win on interface polish, survey depth and the size of its integration library, so the better tool depends on whether you want chat and behaviour in one dashboard or a cleaner research workflow.
What Is the Best Tool for Creating Heatmaps?
For most small and mid-sized sites, Hotjar and Crazy Egg are the two strongest options, with Hotjar better for qualitative research and Crazy Egg better for snapshot-based testing. Microsoft Clarity is the serious free alternative with unlimited sessions, and Mouseflow is worth a look if form analytics and friction scoring are your priority.
Can ChatGPT Create Heatmaps?
ChatGPT cannot generate a website heatmap, because it has no way to record visitor clicks or scroll depth on your pages. It can build a heatmap-style chart from a dataset you upload and it can help you interpret an exported report, but the behavioural capture itself still needs a tracking script from a dedicated tool.
What Happened to Hotjar?
Hotjar was acquired by experience analytics company Contentsquare in 2023 and continues to operate as its own product with its own pricing and free tier. The practical changes since then have been deeper integration with Contentsquare’s enterprise platform and ongoing adjustments to plan limits, so check your session allowance at renewal.
Get a Second Opinion on Your Heatmap Data
If you have snapshots piling up and no clear sense of which finding deserves a test first, SEO Quirk can review the maps alongside your analytics and prioritise the changes most likely to move revenue. Take a look at how we work and send over the pages you want examined.