What do we do with heat maps?
August 2019
Heat maps and scroll maps are great tools for detecting and communicating patterns like “nobody scrolls past this section” or “nobody clicks on this CTA.” How should we act on these observations?
Nobody scrolls past this section
Your scroll map indicates that only 20% of visitors scroll to the bottom of the page.
Is that good? Terrible? How would you even know?
Hotjar gives an example of this type of scrollmap and how they interpreted it. They suggest moving the bottom-most element up on the page.
This might work - more people see it, find it compelling, and click joyfully.
But … maybe people will just stop scrolling higher up the page. Or they’ll scroll right past it without noticing it. Or they’ll see it, find it confusing, and withhold their precious clicks.
You can test moving things up on the page, if it’s worth the effort. But a scroll map doesn’t provide evidence that moving elements on the page will increase conversions.
Here are a couple of actions I’d feel comfortable taking based on scroll map data:
- “Nobody scrolls past the 3rd section on the page” –> Let’s not bother testing elements further down the page
- “Nobody scrolls past the hero section” –> Let’s find out if we even want them to - by testing design that encourages scrolling, and/or swapping out the section immediately below the hero
Nobody clicks on this CTA
When a heat map shows little or no activity on a button, it’s a clear signal that your page layout, or offering, is not ideal.
But since people aren’t clicking the CTA, we can’t know whether having more people click it would be a Good Thing or not. (What will they do after clicking?)
On the same examples page, Hotjar points to a “Play Video” button that used to live in their hero section. Turns out only 3 out of 7,000 visitors clicked it.
Their response was to test redesigns of the home page that removed the video link.
I see the logic, but for all we know those 3 clickers all signed up for Enterprise plans and shared the decision across social media, thus generating more revenue than the other 6,997 non-clickers.
It’s the same problem as with the scroll map; we don’t know the value of the actions being taken or not taken. What’s a video click worth? Would it have been better to test redesigns that make the video link more prominent?
Here are a couple of actions I’d feel comfortable taking based on heat map data:
- “People are clicking the thing we don’t want them to click” –> Okay, remove the thing
- “Nobody clicks on the CTA in the 4th section down” –> Probably do nothing, unless this CTA has somehow become a top priority
- “Tons of people try to click on this non-clickable element” –> Can we make it clickable, and give it an action that’s relevant to conversion? Let’s do that. (If not, can we make it look less clickable?)
I do ❤️ heat and scroll maps as diagnostic and communication tools, but I’m skeptical about taking direct action based on them.
What about you? Do you change your site based on heat/scroll map data? With no regrets? Hit Reply and tell me about it.