Data beyond Heatmaps — that top eCommerce brands track

Ask any retailer what’s one of the tools they like to use for gathering website insights and heatmaps will be a common answer. It’s a widely used and most loved data visualization tool for eCommerce business owners. 

So what are heatmaps really? Heatmaps are visual representations of complex data that in turn represent user behavior: what elements users spend time on, what they completely avoid, and what they get repeatedly drawn to.

In real-time, heatmaps gather data around typical user behavior such as mouse movements, scrolling, eye fixations, etc. 

Heatmaps have been named the way they have been because they color code according to user behavior. While more dense colors are used to indicate more intense activity, less dense colors indicate the opposite.

The hot to cool color spectrum makes complex data come alive — but in an easily understandable way. 

However, most retailers often confuse that heatmaps merely help visualize data and are not DATA themselves! They think heatmaps are the only reasonable way to glean precious user insights. This does not take into account that heatmaps can be erroneously used — thanks to biases resulting from subjective analysis, misreading of data, etc. 

And this is why it’s necessary to use tools and techniques beyond heatmaps to arrive at insights and conclusions around user behavior.

In this piece, we’ll explore 8 factors that you absolutely must consider to understand user behavior better. 

8 factors you need to watch out for “beyond heatmaps”

1. Clicks mean nothing; look for conversions

To enhance the way you are reading into user activity through heatmaps, you have to look at customers that convert.

In this context, clicks can be misleading, because users click for various reasons. Sometimes clicks are a result of users getting attracted to images, titles, and shapes for reasons that don’t match their original purpose. 

In the example below, you’ll notice how people have not only clicked on hyperlinks but also on static elements such as headers, static images, and more.

In that case, observing conversions alongside could be a good reference for you to understand what is motivating them and why they are clicking through to buy or subscribe.

Typically, users who have converted are led to a thank you page. To observe who converts, you can attach a substring to the visited page URL, so that whoever converts can end up on the thank you page. 

Similarly, observing records of users who didn’t convert will offer you further insights. Notice at what point a user fell off, what they were engaging with when the fall-off happened, etc.

To access these findings, you can set up a filter that will potentially contain users who proceed towards checkout but fall off right before.

Interested? Check this out: 5 stages of an ecommerce conversion funnel (+ways to improve each step)

2. Data is king; gather as much as you can through session recordings

What can data do that nothing else can? Well, data can reveal how your users think and have the potential to act.

Data also reveals their behavior so far, offering you clues to make tweaks and changes to your current eCommerce strategy. 

One way to understand users better is to gather as much data on them as you can. Watching session recordings is an effective way to do this.

Connecting an external API to your registered customer database can help you gather data through session recordings. Multiple session recordings across the same user might be able to tell you more about their motivations and complaints.

Sometimes call-to-action buttons are unresponsive and at others, subtle navigational challenges may leave the users frustrated. Watching multiple session recordings can enable you to improve user experience across the different stages of browsing and shopping. 

To view the recordings for every user, all you need to create is a relevant filter. In many instances, a user’s email can be used as a filter.

3. Web forms reveal critical user information; deep-dive into form analytics

One factor that can say a lot about how your users are engaging with your eCommerce brand is how they engage with your web forms. As many eCommerce businesses might think, form analytics isn’t just about understanding who is converting and who is not. 

Through form analytics, you can also potentially find out who is filling the form and who is only viewing it, how many give up filling it up after a point, and how much time on average it takes for people to fill-up the form. Form analytics is also helpful when it comes to analyzing which fields are the most corrected and which aren’t filled out at all.

If you’re using form analytics, you’ll be able to analyze the number of visits to the page that carries the web form, the number that directly interacts with it, and the number that ends up filling up the whole form.

4. Cart abandonment can reveal more than you think; go deeper into funnel analytics

“The cart” represents the end of the user’s conversion journey. And that’s why many eCommerce businesses erroneously think the reason for this is also somewhere around that ‘last step’.

Funnel analytics, however, can reveal that cart abandonment could occur much before —  sometimes on a product page, sometimes on a category page, and sometimes even on a sub-category page. 

When you employ funnel analytics, you will be able to identify factors that may be contributing to bottlenecks within the funnel. For example, you might find out almost every user provides their name and email address but seems to fall off at the captcha stage.

If you find this to be a repeating pattern, you may safely arrive at the conclusion that there is some captcha error. Creating a comparison between the steps you expect will perform well and market benchmarks may be valuable. 

Similarly, how much time users spend between steps on an average may also reveal more about cart abandonment. A clear understanding of the products typically abandoned at the cart level may also offer you deeper insights.

If you notice a certain user segment converting more in comparison to another, it could well be time to change the brand messaging on your website.

Want to win ‘em back? Check out: 20 abandoned cart email examples that actually win back lost customers 

5. No customer is alike; filter your audience

It’s not uncommon to hear the term “average” when it comes to deriving data and then analyzing it. The average rate, the average user, the average tendency...you get the drift.

While the law of averages is great for academic purposes, in reality, every customer and their motivation is different. And when you are employing methods beyond heatmaps, this becomes the primary reason to opt for audience analytics. 

Typically, when you are trying to filter your audience, multiple variables come into play. Here are a number of ways you can create effective audience filters around:

• Navigation:

Filters around navigation will revolve around the point users enter your website, pages they visit along the way, pages they don’t visit at all, and the point where they exit your website.

• Tags and variables:

You can also look at tags and variables to create relevant filters. They can either be built-in or be customized.

• Technology:

You can also filter your audience based on the technology they are primarily using at the time of the session. These could include browser, operating system, device, and even screen resolution.

• Session data:

You could also create filters going by the duration of sessions and how many page views occurred across each. Working with a spectrum for each might be helpful.

• Visitor data:

The location of visitors, how they landed on your website (was it through email, google search, social media, etc.?), and if they are returning or new users, can be used as markers to create filters.

6. Customer experience is varied; understand why friction happens and what causes it

How much friction a user faces during their session experience can be useful data to be analyzed further. The more friction a user faces, the more likely it is that they won’t convert. 

Identifying distinct friction events across session recordings can further enable you to understand user behavior. The following will give you an idea of what to watch out for. 

• Submit failure:

Marking a submit attempt under “failed” can register it as a friction event. This falls under the “high friction” category. 

• Compulsive clicking:

This refers to users clicking on a particular part of the page again and again, within a short period of time. This also falls under the “high friction” category. 

• JavaScript click error:

When a visitor clicks, expecting to be led to a page, but instead a JavaScript error message shows up. This is again a “high friction” event. 

• Bounce back:

This is a “moderate friction” event that occurs when a user is trying to navigate from page A to page B. However, soon after they reach page B, they bounce back to page A.  

• Excess movement:

Considered relevant specifically for mobile devices, this indicates orientation changes and multiple zoom events during a user session. This is considered a “low friction” event. 

• Moving cursor:

This is a friction event that describes the user’s mouse cursor moving away from the page they are on and settling on another tab or program. Another “low friction” event. 

• Speed browsing:

When a user accesses multiple pages within a short period of time within the same session. This also counts as a “low friction” event. 

• Custom friction:

Additionally, you can also create a tag for any event that you have identified to cause friction. 

7. Feedback comes first; get them with customer surveys

While heatmaps offer a relevant window into user behavior, they may not do a really effective job of explaining the “why” behind the “what”. This is where you can count on customer surveys to reveal nuances that you may not have even imagined. 

Micro feedback can really help you get into the gaps and fix your website for a better user experience. So how should you approach customer surveys in order to gather ample feedback?

> Clear and simple questions:

Easy-to-read and easy-to-process questions are more likely to be answered by users than their more complex counterparts. Considering most users just scan through the content they are viewing, complicated questions are likely to be totally skipped over.

Here are a few examples of the kind of questions that work for users. 

• How would you rate your shopping experience?
• Would you tell your friends about shopping with us?
• How soon would you come back to shop with us?

> Smooth, logical flow:

If your idea is to get users to think in a particular direction, it’s best to offer them questions that graduate from one related concept to another. A sketchy arrangement of questions can have a lower success rate. 

> A response scale that’s most relevant:

Choosing a response scale that enhances the goals of your customer survey, is important. Here are the three main kinds of response scales you can choose from.

• Rating response scale:

In this kind, you can create a spectrum with numbers. For example, if you were to ask “on a scale of 1 to 10, how would you rate us?”, it would fall under this category. Businesses most commonly choose a 1-3, 1-5, or a 1-10 rating response scale.

• Dichotomous response scale:

In this kind, a question makes room for either-or answers. True/false, agree/disagree, and yes/no options fall under this category.

• Semantic differential response scale:

In this kind, words are used to create a spectrum so that users can key in their responses.

You can take a look at this example from UNIQLO to see how all these points come together.

Hey, you’ll love this: 15 EASY ways to get customer reviews and boost sales 

8. User investigation can reveal nuanced data; bring usability tests into the mix

Along with all the quantitative data you collect, qualitative data can prove to be your real-time key to an improved browsing experience. Usability tests can deepen your understanding of motivations and challenges.

According to the Nielsen Norman Group, 5 aspects decide how usable a user interface is: 

• Learnability:

This defines how easy it is for users to engage with the interface for the first time. 

• Efficiency:

This decides how easy it is for users to apply what they have learned to complete a task. 

• Memorability:

This is a marker of how easily a user is able to re-engage with the interface after a period of not using it.

• Errors:

This can help you understand how many errors users are making, how severe the errors are, and how easy they find it to get back on track. 

• Satisfaction:

Finally, the marker that decides the overall satisfaction level for users. 

In usability testing, you typically set users up for a task and then observe how they go about completing this task. You can observe live sessions and also create follow-up sessions to understand their takeaways about the overall experience.

Heatmap FAQs you’re better off knowing

1. What are website heatmaps?

Website heatmaps are visual representations of critical user data. They are data visualization tools that throw light on how pages on a website are performing.

They typically come with a warm to cool color scheme — the warmest color representing the highest user engagement and the coldest color representing the lowest.

You’ll notice how this example from Forever 21 has the warmest color density on their most popular products, showing them what captures most attention on the product listing page.

Commonly, 5 types of website heatmaps are used: click maps, scroll maps, eye-tracking heatmaps, mouse tracking heatmaps, and AI-driven attention heatmaps.

2. When should you use a heatmap?

The answer to this would be — as often as you can!

The reason behind this is simple. While quantitative research tools like Google Analytics record the number of users who visited your website, how many converted, and how many bounced, qualitative research reveals more.

Heatmaps are qualitative research tools that help you understand:

• Why certain visitors are spending more time on your website
• Why certain visitors are converting as opposed to certain others
• Why the bounce rate is higher for some pages than others

However, if you’re trying to narrow down on the use of heatmaps to achieve certain specific goals, then the following will help you decide.

• Introducing more intuitive UX and UI:

Heatmaps can help you zero in on what will make the elements on your website more visitor-friendly. Using heatmaps can definitely help you bring changes to design elements, CTAs, choice of text as well as how it is all placed.

• Reducing bounce rate:

You could use heatmaps to understand why visitors are spending less time on your website or why they are not converting. Heatmaps help businesses zero in on potential distractions, errors, and bottlenecks.

• Increasing the number of leads & sign-ups:

Heatmaps can help you do this by offering insights on forms, their placement, and how they can be optimized. 

• Improving the quality of navigation:

Have heatmaps come to your aid to offer insight into your current navigation (what is working and what is not). This can potentially help you reduce friction and improve the overall user experience.

• Running A/B tests:

Heatmaps are a great precursor to A/B tests because they serve up enough data. With this data then, A/B tests can be run to gather further understanding about metrics and impact. 

3. How can heatmaps be misleading? 

Heatmaps visually capture data and one can potentially make the mistake of reading this data. For example, just because a visitor clicked on “Sign Up” does not automatically mean they finally registered as a user.

Heatmaps don’t account for the likelihood that the user may have received an error message saying the email they are trying to sign up with is already registered.

Similarly, heatmaps may not be able to help in understanding why a user behaves in a contradictory way and clicks on paradoxical elements featured on a page.

4. What are the limitations of a heatmap?

• Ignoring responsiveness:

Heatmaps work best with static pages and layouts. Unfortunately, we live in a time of varied devices and responsive design. Even if you select a specific area to be studied on a heatmap tool, in reality, this area will represent different things across different devices and resolutions. 

• Not factoring in dynamic websites:

Heatmaps are not really functional with dynamic websites. And if you think about it, that’s how most websites are these days. From animation to slide-out panels to modal dialogs, heatmaps aren’t able to keep up with page changes. 

• Relying solely on heatmaps:

While simple visualizations make heatmaps very popular, they are not all that helpful. Using only heatmaps without any other tool could mean acting on the wrong hunches or not being able to cross-reference data. 

• Failing to take sample size or time periods into consideration:

Heatmaps represent data, yes, but they are not fine-tuned enough to capture the size of a data set. If you forget to adjust the settings so that the sample size reflects the number of visitors in real-time, the data you’ll be looking at would be wrong. 

• Not segmenting new vs returning visitors:

Many marketers make the mistake of lumping all visitors together within a heatmap. This is a problem because new visitors and returning visitors behave very differently from one another. 

• Not aligning heatmap analysis with customer feedback:

Heatmaps are a great tool to analyze how customers are reacting to different parts of your website and even different parts of the same webpage. However, it cannot replicate customer feedback. Along with heatmaps, only if you set up certain relevant customer feedback mechanisms, you’ll know how to make the right changes.

The following feedback tools can be helpful:

• Feedback widget:

This appears as a button, which when clicked allows the user to highlight specific parts of the page they are visiting. Right after, the user is also able to offer feedback for the highlighted parts. 

• On-page survey:

This appears like a chat window and can throw up results for a number of markers - time spent on a session, depth of scrolling, number of pages visited in a session, and likely intent for exiting, amongst others. 

• Exit popup survey:

This appears like a quick survey just as a user is about to jump off a specific session. The questions are typically short, very relevant to a user’s quality of experience, and come with multi-choice answer options. 

Here’s a very simple example of an exit popup survey. Even something like this does the trick.

• Fast conclusions:

It’s easy for marketers to get excited over results without observing the trends across the whole heatmap sample period. This can lead to assumptions and misreadings, which can further be used to make the wrong website changes.

Want to continue reading? Check out: 31 eCommerce metrics & KPIs that actually matter