User research methods: A guide for UX designers

If you were buying a house, you wouldn’t just walk into an estate agent’s office and purchase whatever your gut tells you to. After all, you’re probably spending a lot of money on this purchase and likely don’t have the money to buy another (…or another) if it doesn’t meet your needs.

Fixing problems with the house after purchasing it often costs an eye-watering amount, too. The cheapest option, keeping in mind that the only “free” option is picking a house and praying that it just happens to be a winner, is spending a little time and money on research.

This idea applies to everything, including UX design. Investing a relatively small amount of time and money on user research, sometimes referred to as UX research, could reduce or eliminate the risk of designing an app or website that doesn’t convert users. Without some initial user research, you might have to spend a larger amount on fixing (or patching) the problems or start over completely.

What is user research?

User research is the process of learning what users need and want from an app or website. This enables UX designers to create a positive experience for users, usually with the added benefit of more revenue.

Formative user research methods

We’ll start with formative user research methods, meaning any methods that are used before designing anything. Formative user research is all about discovering, validating, and understanding problems. This is sometimes referred to as being in the problem space.

Surveys

As a method of formative user research, surveys are the best overall. Why? They’re unmoderated, easy to implement, and flexible in terms of when and how we’re able to present them to users. While they’re not the be-all and end-all of user research, they’re ultimately very useful while being very cheap.

Use surveys when you have questions to ask or answers to monitor. Be mindful of what you’re asking, as well as when and how you’re asking it. Here are some survey types that you don’t want to miss out on:

• Email survey — Acquire feedback from converted users
• 24/7 survey — An always-visible in-app or on-site survey that draws feedback from your happiest and angriest users, due to there being no pressure to take it
• Post-action surveys — Identify problems with important functions while the user’s thoughts and feelings are still fresh (targets all satisfaction levels, applies moderate pressure)
• Exit surveys — Focus on what’s making users leave (targets all satisfaction levels, applies moderate pressure)

Interviews

Interviews are basically one-on-one surveys. They typically happen in real-time and come with the option to be more reactive to answers. However, interviews cost more time and money than surveys do, so make sure that you screen participants carefully.

Field studies

Field studies are exercises where users are observed before, during, and after using an app or website. These exercises enable UX designers to focus more on the context of users’ problems and how they go about solving them.

You can do field studies with your own app or website, or with a rival’s. This study is more akin to writing a biography than recording survey responses, although you will be asking questions.

Diary studies

Essentially, diary studies are unmoderated field studies. Instead of being observed, users keep a diary of their actions, thoughts, and feelings before, during, and after using an app or website.

Analytics

Analytics tools collect user data such as demographic, behavior, and acquisition data. UX designers can use this data to discover potential problems with an app or website, potentially pinpoint the source of those problems, and collect potentially contextual information.

A drawback of analytics tools is that you can’t validate any of the information they collect, which is why I’ve used the word ‘potentially’ a lot. Analytics tools are great and sometimes free, but the raw data doesn’t become useful information unless it’s used to supplement other methods of user research such as surveys and A/B tests.

For example, imagine a user of a certain demographic, acquired via search, landing on an article, progressing into a conversion funnel, but then dropping off at the cart. This would be quite a lot of data, and yet it still wouldn’t confirm why they dropped off. In fact, we’d be lucky that they dropped off at the cart as that heavily suggests intent.

However, combining this data with survey results could produce a more complete picture of the user and their problem. That being said, if you believe that the data alludes to a problem with a very simple solution, you could take a chance on a low-risk A/B test.

Ultimately, user research isn’t a linear process. You have to weigh up the cost of research and development and decide whether the potential resulting clue or answer is worth the risk, and if not, then consider taking a different route with different user research methods. More on that later, though.

Heatmaps

There are two types of heatmaps — clickmaps and scrollmaps. Clickmaps reveal what users click on and scrollmaps reveal where they scroll to.

While “rage clicks” can confirm visual affordance problems and buggy interactive elements, heatmaps are otherwise just another type of analytics, and should be treated as such.

Participatory design

Participatory design involves users sketching their ideal design. You can use these sketches to uncover solutions, but also problems in the case where the sketches don’t match what your design looks like already.

Open card sorting

Open card sorting is an exercise where users label and sort navigation categories and navigation items in a way that makes sense to them. This enables UX designers to design effective navigation components and flows. Follow up with closed card sorting to validate what you ultimately decide on.

Summative user research methods

Now let’s explore summative user research methods, meaning any methods that are used to validate designs or decisions that you made based on what you learned from the formative user research methods. This is sometimes referred to as being in the solution space.

Closed card sorting

Closed card sorting is the summative version of open card sorting. With closed card sorting, the navigation categories and items are already labeled, and users simply need to clarify which category they’d like each item to be in.

Tree testing

Tree testing is an exercise where users are asked to find navigation items within an unstyled navigation. UX designers can use tree testing to identify labeling or categorization issues, formatively leading into open card sorting if necessary. You can also use tree testing to summatively validate the outcome of a closed card sort study, which is more common.

A/B testing

A/B testing, in the context of UX design, is a controlled experiment where a new version of a design is compared to an existing version in order to measure which one is objectively better. To ensure a fair experiment, it’s best to change only one thing at a time.

This approach works best for businesses with many users or high revenue where these small changes can have a massive effect. It’s also low-risk, so you don’t necessarily need to lead into it with any formative research.

When testing multiple changes, for example when testing a function, take more care and do formative user research first, because although A/B testing is unmoderated and you won’t need to recruit participants, developing the changes costs time and money, even if they aren’t used in the end.

Multivariate testing

Multivariate testing helps UX designers to test multiple combinations of changes at once, solving the problems that come with A/B testing. It can be quite complicated though, and you’ll need a large sample size — possibly a huge sample size, depending on how many changes you’d like to test.

Usability testing

Usability testing is an exercise where users are asked to complete specific tasks, leading to the discovery of all kinds of user problems. UX designers can also use usability testing to benchmark improvements over time.

Usability testing tools can capture the user’s screen, voice, and face. They also tend to have analytic, heatmap, survey, and A/B testing features, although not multivariate testing. These tools enable UX designers to acquire an insane amount of data, insights, and context when testing apps, websites, or prototypes of either.

Usability testing can be moderated or unmoderated.

The user research process

There isn’t a specific way to do user research. Bandwidth, budget, business objectives, known business problems, and risk level are just some of the things that can impact the route you take on your user research journey.

Ultimately, just stay mindful of your business objectives, what you need to learn in order to advance those business objectives, and then which user research method would be best for turning over the next stone, so to speak.

Working backwards like this helps you to avoid doing user research just for the sake of it and also focusing on vanity insights that aren’t useful. This is the strategic approach.

If you’re not sure where to start or nervous about doing research, take the safe approach instead. Install 24/7 surveys and exit surveys onto your app or website, or if you don’t have an app or website up and running yet, steal the participants that your product team has been using for product research and send them an email survey.

A very basic “What’s the worst thing about our app/website/prototype?” is enough to start off with. Then, once you have enough data, make sure that you effectively synthesize and validate it. If you’re not sure whether you have enough data, collect more.

To summarize

As we’ve explored, there are two types of user research. Formative user research methods are more about problems, whereas summative methods are more about solutions.

Other than that, every user research method tells a different part of the story. So, while you can certainly strategize in terms of roadmapping and budgets, you’ll never really know how your user research journey will go and even what user research methods you’ll end up using. There’s also no single, specific “right” way to do user research.

However, there’s one thing that we do know for sure and it’s that user research is the safest way to design products. It’s often the cheapest and fastest way too. This is because making design decisions on a whim — not including design heuristics, of course — often results in products with a poor user experience.

Got a question? Put it in the comment section below!