How to Identify and Fix a Broken UX with User Behavior Analytics

Lorraine Ryshin

December 6, 2019 | 6 min read

Last Updated: Aug 14, 2020


Some website users undergo a bad UX, which leads them to exit — or worse — bounce from a website, possibly to never again return. Understanding what causes premature site exits is key to improving the customer experience (CX), and delivering journeys that help customers meet their wide-ranging digital expectations.

Making use of data for a UX analysis is the most practical approach to scrutinizing customer journeys, including high-level views that locate friction points and counter-intuitive navigation patterns. Once you’ve identified your problematic pages through a high-level view of user behavior, you can make more fine-tuned changes by assessing individual pages and elements.

Achieving a fulfilling digital experience is attainable, but you have to identify what constitutes a broken UX in the first place, and establish the visitor segments that come across one. Once you have this insight on hand, you can prioritize optimization efforts to improve your digital experience and make your visitors crave more.

Identifying What’s Amiss in the Customer Journey

We quizzed Ying Yang, our Lead Product Experience Manager, to get her thoughts on where to start. “The first thing you must look at when identifying a poor UX is the customer journey,” she said. “You should be able to break it apart page by page to see exactly how users traverse your site during each session.”

A well-built customer journey analysis tool will show you each step a customer takes during their time spent on a site, help uncover what they are trying to do, and how they went about doing it. You ought to be able to detect where the first UX friction lies on a high level; to find this, you have to pinpoint where users are bouncing or leaving the site, and what led to this outcome.

“You need to identify the last page that a segment of users stayed on during their journey before leaving your site. It is this page in which their UX was disrupted,” explained Ying. 

However, in longer customer journeys, note that a page from which a user has left the site may not signify a bad experience. Instead, the user may simply feel that their stay on the site is complete, and requires no further browsing.”

As such, observe the pages that contain bounces initially, as there is some shortage of retaining the visitors’ interest. Furthermore, since a bounce is more caustic than a regular site leave, it requires immediate attention. (Bounces reveal a non-existent journey, or one of one step/page visit).

Now that you’ve found the page with the UX culprit of bouncing or exiting, let’s delve further. 

A Further Analysis of a Crippled UX

Entering step two of making corrections, you will need to work out the cause behind particular site exits or other behaviors indicative of frustration or unmet needs. In order to spot individual obstacles in the customer journey, you’ll need to analyze specific elements within a page. 

Through this approach, you’ll be able to catch the exact cause of friction (whether it’s a CTA, image, product description, form field, etc), as opposed to guessing what regions and elements of a page led users to leave.

So what do you do when analyzing a particular page element? You take a hyper-focused turn in your UX analysis. “This is a more granular step,” says Ying. “As such, you’ll want to look at a robust batch of behavior and revenue metrics. These present a deeper dive of your UX to follow up the customer journey analysis.” 

Here are just a few of the metrics you can appraise for a granular UX performance check:

Hover Rate: The percentage of pageviews in which visitors hovered over the zone at least once, determining which zones are consumed the most. This helps you rank zones and assess if they are consulted properly, by weighing in factors like averages of other zones and the page length. 

Click Recurrence: represents the average number of times a zone was clicked when engaged with during a pageview. This exposes either engagement or frustration. For example, a high click recurrence on a carousel is good news, as it shows a high engagement with an element offering many clickable areas.

It can also point to frustration. For example, if users click on the same element multiple times — such as an image or link, it means the element is drawing up errors; it’s either unclickable or not performing its function correctly. 

Conversion Rate Per Click: Applying only to clickable zone, this metric relays if clicking on a zone impacts the user’s behavior or conversion goal.This helps you determine which elements contribute to or deter from conversions. A conversion can be any behavior you set. 

Exposure Rate: identifies how far down a page a user scrolls; it’s accounted for when at least half of a zone is viewed. This helps you understand how much users scroll, allowing you to make empirical sizing adjustments.

Attractiveness Rate: Relays the percentage of visitors who clicked on a zone after having been exposed to it. This informs you on optimizing the placement of content on your page. For example, if more users click below the fold, you should move that content further up for more of them to see it quicker. A high rate proves the high performing attractiveness of an element.

Segmenting Your Users for UX Comparisons

After you analyzed the elements of your page with granular behavior metrics, you’ll need to analyze further, by conducting comparisons. This will help you determine what comprises an underperforming UX more clearly. To do this, you would need to compare a good behavior with a bad behavior.

Comparing the experience of visitors who accomplished the goal of a page with those who didn’t, will further confirm what needs fixing. You can carry out a zoning analysis on these two segments as well as make comparisons on each metric. 

This allows you to catch where non-converting visitors tend to hover and where they are more inactive. But most importantly, it allows you to weigh this data against the users who did convert/ achieve what they came to your site to do. 

“For example, you can build a segment for the users who saw a 404 error page and compare it with the ones who had the same issue across different journeys or those who didn’t run into it,” explained Ying. “Additionally, you can create a segment around users who clicked on a CTA, deepening their journey against a segment of users who didn’t, or worse, ended their journey on that page.”

Main Examples of UX That Cuts the Customer Journey

One of the attributes of a broken UX is content that doesn’t engage users or is not seen, thus prompting visitors to exit the site. Pages that require too much scrolling, for example, may yield low engagement or little to no views.

For example, a particularly wide banner that takes up much of the screen may be obscuring other content that’s crucial to generating revenue. Some users may not even be aware of the content below the fold. 

“Most high-performing content should have real estate above the fold,” Ying advises. “Does your business have a major campaign or sub campaign running? Post more than one type of content about it above the fold. These can exist as tiles, a carousel or both.”

This source of friction is especially damaging to mobile UX, which has a much smaller screen size than desktop. As such, some functionalities aren’t well suited to be crammed in. “Big banners, images and accordions (vertical menus) push everything down below the fold, so don’t overuse them. You will probably need to scale back on some of these elements to avoid a UX that has turned sour.” 

Another example of poor content occurs when banner usage is slight and/or doesn’t achieve the goal of a page. For example, a banner can send users to a PDP (product details page) that cuts off their browsing journey.

“PDPs, in general, have high bounce rates, as in the case of our retail clients, so you need to be careful what products you send users to, should your banner send them to a PDP (or even a product landing page). Landing on a PDP is especially detrimental to the user experience when the real goal was to send users to a PLP (product landing page), which shows several product options as opposed to a PDP.”

Adobe Stock, via studiostok


Fixing Customer Journeys

Now you know how to move the needle from a high-level UX analysis to a granular level to spot what caused your customers to struggle or give up with your site. After you identify what leads to bad digital experiences, you are all set to start optimizing. Customer experience analytics are your best friend when it comes to augmenting your content ideation strategy.

Since it allows you to meticulously identify digital experience issues, it fastracks you to brainstorming sessions to rectify the issues in a data-backed way. Some things will be clearer than others. For example, if you find 404 errors and other dead-end pages, the quick fix is the get rid of them, or replace them with the proper pages. 

“For example, if an item is no longer in stock, or no longer being digitally offered, make sure it doesn’t yield the 404 error. But if it’s a product users can purchase, or if a page offers any other type of conversion (signing up for content, etc.), make sure your page is functional and devoid of any confusing elements,” said Ying.”

 

Hero image via Adobe Stock, by Marvi7