Skyrocket Conversions By Fixing Your User Experience

By | 23rd November 2017

A broken user experience can completely kill conversions and yet many companies make changes without evidence and launch features no one wants. Why do companies put so little effort into fixing their user experience? Could it be that no one ever gets promoted for fixing a broken user experience, but shiny new features are much more likely to be noticed by senior management?

But as conversion rate optimisation expert Craig Sullivan  pointed out at the Conversion Elite  conference, fixing broken stuff and improving the usability of an experience is much more likely to improve conversions and revenues than using persuasive techniques or increasing the motivation of users. So, how should we approach online optimisation?

 

How Do You Fix Your User Experience?

Before even considering running A/B tests we should be answering these questions about our digital user experience. If we examine these issues first we are likely to identify a majority of the problems that probably cause customers to give up and go elsewhere.

What Do We Already Know?

At this point there is a strong temptation to ask your conversion rate optimisation team to start browsing your digital experience to look for broken stuff, but this is time consuming and wasteful as you probably already have access to data that indicates where problems may lie. Interrogate your web analytics (including error rates), form analytics, user research (e.g. session replays), surveys and feedback (e.g. complaints) to observe real journeys and identify touch-points that indicate where there may be problems with your user experience.

Hotjar is a very cost effective user experience solution which includes session replays, click heat maps, form analytics and user feedback. Make sure you schedule time to view session replays as this is an often underused feature that allows you to observe real user journeys. There are a number of cost effective  user experience tools to choose from if you don’t already have one for your site.

It is also important to analyse your data by device class (mobile, tablet and desktop), but also operating system and screen resolution. For responsive site you should also look at breakpoints to see how this impacts upon conversion. But also don’t forget to try to track people rather than devices. Google Analytics and other analytics solutions allows for a user view to monitor logged in customers across device and so you can measure the entire user journey and calculate a true conversion rate.

What are the use cases?

Context is everything as it will often determine the current user need. For example if you are using a train ticket app your need will be very different a week before you travel compared to when you are at the station barrier or on the train with a ticket inspector asking to see your ticket. The user may also have to rely on 4G rather than WI-FI when on a train and so turn off your WI-FI to observe the real experience when testing an app.

 

How do you test your own digital experience?

Now you have data on real user journeys it’s time to immerse yourself in your digital experience. Using the data you collected you can focus on the devices, browsers and journeys that customers follow. You won’t have the time or devices to check ever user experience and so use a cross-browser testing solution to get screen shots of all the main devices and browsers you need.

However, you should still get your core user devices so that you can get a better understanding of the real user experience. For touch devices don’t be lazy and use a desktop simulator as that won’t allow you to observe a touch experience or test usability.

Here is an example of a error-ridden form from Iceland Foods (groceries.iceland.co.uk)  which displays persistent errors for date of birth and the alternative phone number field. These kinds of errors often occur because of lazy programming. Don’t force users to enter data according to your perception of the “right” way to submit data. Instead get programmers to use rules to convert data to a consistent format. Further, why does any company need two telephone numbers, especially when increasing numbers of  people only have a mobile number? See more web form design best practices here.

Image of Iceland Foods sign-up form with With errors on DOB and alternative telephone

Does it load and work quickly?

You can measure actual page speed using Google Analytics or other web analytics solutions. Given that many sites will continue to call on the server for content long after the page has initially loaded use the Average Document Interactive Time in GA. This is the time that it takes for the content to load sufficiently for a user to be able to interact with the page. See more on load speed tools in this blog here.

 

Can people use it?

If your site or app is prone to bugs and errors this can be very frustrating for users and can destroy your credibility. Here is an error that occurs in the popular Meetup.com app that prevents the user completing a RSVP. This kind of bug is very annoying for users as it prevents them completing an important task. You should monitor error rates in your analytics as they are a clear sign that something is broken.

Image of error causes by bug in meetup.com app

User research is also an important tool to find where visitors are experiencing problems. Usability research involves recruiting people who roughly fit your target audience and giving them a simple task based upon a relevant use case (e.g. open a new account or find a certain product on a website). Where possible these should be ‘real users’ rather than employees of your organisation. So, what are the options for conducting UX research?

Face-to-face research:

The advantage of getting someone to navigate a user journey whilst you sit next to them is that you can observe their body language and their level of concentration to better understand how they react to the new design. This can improve the richness of the feedback you obtain from each respondent and it may allow you to ask more relevant questions as you observe their behaviour in its entirety.

On the downside people can behave differently when they are aware they are being observed and so you should allow time for the respondent to settle into the process. You can also reduce this effect by including the target experience in a series of tasks that hides which site you are testing. You might for example tell the participant you are testing a competitor site and then ask them to go to the target site for a comparison.

Face-to-face user research doesn’t have to be expensive. If you don’t have a budget why not go to your local coffee shop or bar and offer potential participants a free drink if they agree to take part. This will take you out of the corporate bubble and you should get a variety of users to observe and learn from.

However, always use appropriate software to record your research sessions as you can’t concentrate on everything going on and ask relevant questions. Check out UX recording software, like CamStudio and other solutions listed in my Digital Marketing Toolbox.

Remote User Research:

Conducting moderated user research remotely can be relatively cheap depending upon whether you are doing it yourself (e.g. using Skype to share screens) or using a UX agency. It also allows you to recruit people from any location in the world if you need to and so you can allow for cultural differences in behaviour and ensure you get a diverse set of participants.

Check out this comprehensive guide on remote user research from @anatinge and you can access a list of software solutions in our Digital Marketing Toolbox.

The advantage of remote moderated research is that you have more control over the process and if necessary you can guide participants to the next step in the user journey. Bear in mind that any intervention will bias the response, but provided you understand this it can be useful to ensure users test each step in a journey even if you know they needed some help to get through it. You can also use moderated research to help guide the development of an automated user research study.

Remote Automated User Research:

These tools generally allow you recruit your own users or they will find people for you from large panels of people that regularly participate in user research. If you are using their panel you can specify recruitment criteria to ensure they roughly meet the demographics of your target audience. However, don’t get fixated with recruiting a representative sample of users as this is not appropriate for user research where you are testing the interface, not the product’s appeal.

Once you have outlined the task and if necessary provided any screenshots (e.g. for a new design that is not yet live) these tools can provide recordings of the task or scenario based test results back very quickly. Check out my post on automated usability testing here.

Quick Feedback:

There are a number of tools that offer quick feedback on concepts, such as the Five Second Test from Usertesting.com and Usabilla.com. I’ve summarised the top remote usability solutions in the blog How to do usability testing to improve conversions which will guide you how to get the most out of these tools.

Information Hierarchy:

For an evaluation of how easy it is for people to find information or merchandise on your site check out tree testing tools here and for an assessment of your navigation categories see this post on card sorting tools. The information hierarchy is an important element of the user experience that is often ignored as it’s taken for granted, but could be a reason for a sub-optimal conversion rate.

Voice of Customer:

Don’t expect users to be proactive and tell you what is broken on your site – they won’t as you are not their priority. But voice of customer tools, including surveys and exit polls can encourage users to provide valuable feedback on the user experience. Check out how to use VoC tools to improve conversion and a summary of over 20 popular tools in this blog post.

Can People Read It?

Accessibility, which includes the readability of content, is an often over looked issues. However, 1 in 12 men and 1 in 200 women suffer from colour blindness. One in seven people in the UK have some kind of disability and 6.1m users in the UK have impairments that affect the way they use the internet.

Usability testing should pick up some of the accessibility issues with your design, but you can also try to conform to the Web Content Accessibility Guidelines (WCAG 2.0). To test if you site has accessibility issues you can use the free Chrome plugins:

Accessibility – Siteimprove Accessibility Checker

Vision – NoCoffee vision simulator

Colour selection – ColourPick Eyedropper.

Screen reader – ChromeVox

 

Conclusion:

A lot of time and money is put into marketing to persuade and motivate people to complete conversion actions. However, as we have seen much of this money is often wasted because simple faults are not resolved. There is little point in pumping money into sending more traffic to your site if it doesn’t work on the user’s device, it takes too long to load, usability issues have not been fixed or if people struggle to read the content.

This often happens because organisations are obsessed with launching the latest new shiny feature rather than getting the basics right. However, as Craig Sullivan pointed out at the Conversion Elite conference, the benefits of fixing just a few of these problems can often be greater than the entire annual IT budget. So, even if it is boring, it is definitely worth the effort.

 

You can view my full Digital Marketing and Optimization Toolbox here.

To browse links to all my posts on one page please click here.

  • About the author:  Neal provides digital marketing optimisation consultancy services and has worked for  brands such as Deezer.com, Foxybingo.com, Very.co.uk, partypoker.com and Bgo.com. He uses a variety of techniques, including web analytics, personas, customer journey analysis and customer feedback to improve a website’s conversion rate.

Leave a Reply

Your email address will not be published. Required fields are marked *