Loading Jaywing website
2 May 2019 / Opinion

5 things to remember when conducting user research

Alice Clayton

When it comes to designing and optimising your website, user experience is key. This means that user research should underpin everything you do.

It helps to give a more realistic understanding of user behaviours and prioritise actions based on what actually impacts user experience.

Moderated user testing is the ideal method, particularly in conjunction with eye-tracking technology but remote testing is still useful, especially for smaller budgets.

Moderated tests allow you to dig deeper into how users are getting on and helps you to understand the ‘why’ behind the data.

To get the most from your user research, it’s important to remember the following tips:


1. Pre-screening questions are key

It’s important to run tests on users who are as close to your target audience as possible.

Ideally, you should test actual users of your site but, if you can’t, you need to be as accurate as possible when recruiting participants so use demographic filtering to get as close to your average user as possible.

When conducting remote user research, use pre-screening questions that identify the users who are most likely to use your product or service.

For example, we conducted tests for a client that provides holiday attractions in Florida.

A screening question asked users whether they had booked a holiday to Florida, or were planning to do so.

It’s important to include dummy answers to avoid users selecting the answer they think will get them into the test. So, for our Florida client, we also included Mexico, France and Spain as holiday options.

2. Have specific tasks to generate actionable insight

If you don’t set specific tasks, users will just aimlessly browse around the site, which might not provide you with useful insight.

You can use insight gathered from heuristic reviews and tools such as Google Analytics (GA)/Hotjar to identify areas for further opportunity.

It’s possible that heuristic reviews identify that navigation on your site is difficult to use or maybe GA showed a drop off during the checkout process, or Hotjar screen recordings show users struggling with a certain feature.

For example, a heuristic review on a hotel site identified that key information about the rooms was hidden, so users were tasked with finding a room with specific requirements, to see if users struggled with this hidden information.

The tests revealed that they did.

Ensure that some of your tasks also address both business needs and user needs. Business needs might involve completing the checkout process to find out if there are roadblocks to conversion.

User needs might be how easy it is to use the website to find what they need.

Giving users tasks structured around identified issues, business needs and user needs should generate insight that can address these issues.

3. Have specific goals for users to accomplish

When setting tasks, it’s important to make sure there’s a specific goal for the user to achieve. Vague, open-ended tasks, such as finding out some information about X car type, are open to interpretation by users.

Some users might whizz through the tasks, whereas others might take ten minutes or more.

They will have different definitions on when the task has been achieved or they might struggle to know when the task is finished.

Although you can still get valuable information from open-ended tasks, the quality of this insight will vary by user.

Having specific goals such as, find a car that meets X requirements, means all users should have the same end-point, will know when they’ve finished the task and you can see if there are any major issues stopping them from accomplishing the task.

This should give you the actionable insight you need.

4. Don’t assume users will use your website in the same way you would

You spend a huge amount of time on your website but your average user doesn't.

When setting tasks and analysing results, don’t expect users to complete a task in a specific way because they don’t know your website in the same way as you.

If your tasks are dependent on users navigating the website in a particular way, prepare for many users to fail the task.

However, it’s very valuable to see how actual users interact with your site, rather than how you expect them to. It could indicate the need for further testing, such as testing the navigation via tree testing or card sorting.

Users will also behave differently to each other too, completing the same task in a different way.

This shows there’s often no one way to complete a goal and forcing users to do so in a way that’s unnatural to them could lead to them abandoning your site without converting.

5. Don’t just listen to what users say, watch what they do too

What users say is different to what they do. They might claim to have found a task easy but you might have just watched them struggle with it for five minutes. Or they might be saying what they think you want to hear.

So, don’t just listen to what they’re saying. They will have lots of valuable information for you but you should watch what they’re doing too.

Are they clicking into lots of pages to try and find something, suggesting navigation is unclear? Are they scrolling up and down the page looking for certain content?

Eye tracking with moderated user research adds another layer to this. It allows you to see what they’re looking at on the page, what catches their attention or what they get stuck on, even if they’re not talking about it.


To be successful with your website and encourage conversions, user research is essential.

It can be costly but, using these tips will help you to make your user research much more effective, both in setting the tasks and analysing the results.

Once you have your user insight, make sure you create a plan for how you will address the issues you have found.