User interview analysis: Steps to turn data into decisions

It’s the moment your UX team has been waiting for. You’ve conducted a series of user interviews that dove into the problems and priorities of your users—now it’s time to dissect the qualitative data and turn it into decision-driving insights.

Keep reading for our chapter on all-things user interviews analysis, where we’ll cover key steps to analyze your user interviews, best practices to center your process around, and what’s next after you extract these insights.

When we talk about analyzing user interviews, more often than not we’re talking about some form of thematic analysis. This method of qualitative research analysis helps you identify themes and patterns in user interview data that you can then use to inform product decisions and build user-centered features, products, and experiences.

Thematic analysis can be as complex as you want it to be. You can closely follow all six of these steps, or loosely convert your data to themes and insights.

At its least intensive level, thematic analysis looks like skim-reading your data and taking note of recurring themes and thoughts. On a deeper scale, you focus in and meticulously assign codes to each relevant piece of feedback—this enables you to truly understand the themes across your data and their relevance to the overall research study.

Thematic analysis can be done manually, as we explain below, or can be done with the assistance of a user interview analysis tool. Both are productive in terms of gleaning valuable insights, but one’s a lot quicker than the other (no prizes for guessing which).

Effective analysis requires you to never lose sight of your original user research plan and goals. By touching base with what you set out to achieve by conducting user interviews, you’re priming yourself to look for relevant patterns, themes, and insights.

These research objectives should act as your north star throughout the interview process, and analysis is no different. Once you’re clear on what you’re aiming for, you’re ready to dive into your dataset.

To analyze your data, you need to convert it to written word. After your user interviews, you’ll likely have hours of audio and/or video recordings.

You can transcribe this manually, by listening to the recording and typing out the responses, or automatically, by using a transcription tool like Maze Interview Studies or a transcription-only tool.

You’ll want to organize your scripts for easy access, helping you cross-reference participants and their answers. There are two ways to do this:

Once you have your user interview scripts, it’s time to perform the actual interview analysis.

Begin thematic analysis of user interviews by familiarizing yourself with the data and reading through participant responses multiple times.

Note any responses with terms or phrases relevant to your UX research goals. You’ll want to return to these notes later, so mark them down by assigning a code.

A code should summarize your point of interest in a few words. Let’s say you run into the following response:

“I have trouble navigating my profile. There are too many options. The worst part is going through them all, I can’t find a search bar anywhere.”

A good code for the first bolded phrase would be: ‘navigation issues’, while the second code could be ‘overwhelming profile features’. The third code could read: ‘search function findability’.

Your assigned codes will be the basis for further organizing your data into themes, and usable insights.

When looking through your user interview scripts, also consider any notes you or a colleague took during the interview, as this can help get deeper insights. Scott Hurff, CPO and Co-Founder at Churnkey, suggests asking yourself:

User interview notes can support your analysis, so keep them close to hand throughout the process.

While reading through your data and assigning codes, you’ll notice relevant patterns, sentiments, and themes starting to emerge. Once you’ve sifted through all your responses, note down your codes on a different document and group them into overarching themes.

Scott explains his approach to this step:

“Once common themes begin to emerge, I like to group them in a spreadsheet and chart them to get a sense of volume. What themes emerge the most? Which type of customer is experiencing this (for example, small businesses vs. indie hackers vs. mid-market businesses)?”

Your chosen theme names should label each code’s subject matter, making it easy to find relevant insights and points of interest. Theme examples could include:

If you’re conducting thematic analysis manually, try sorting codes into temporarily flexible themes using a whiteboard or sticky notes. This makes it easy to move themes and codes around until you find a place that fits each best. Expect to move things around as you assign codes and more themes emerge.

Once your data is distilled into broad, overarching themes, your insights are ready for the taking.

At this stage in the thematic analysis process, you’ll start to uncover customer insights directly connected to your user research goals. Look over the reocurring codes that appear across a single theme to find out what’s most pressing for users.

If many users are stating they “can’t find the search bar” under the ‘search function findability’ theme, and your UX research goal is to improve navigation—you’ve just found an actionable insight.

Highlight responses from recurring codes, and keep on combing the rest of your themes for similar ideas or codes to uncover relevant pain points, usability issues, or particular product impressions and sentiments.

This is also when you might find it beneficial to restructure the themes that have emerged. Don’t be afraid to regroup and divide your themes into a new order—it’s easy to miss connections when first identifying themes, or to identify better structuring and theming opportunities for codes as your analysis evolves. Take another look over your themes to find the best way to organize them.

Once you’ve defined your high-level themes, you can use them to inform the next stage of the UX design and development process. The specifics depend on your goal, but use the themes you’ve identified alongside other forms of UX research to outline a path toward your desired outcome.

The final step in analyzing user interview data is to share your findings. You’ll want to create a UX research report for key stakeholders—such as your team, other departments, and higher-ups—to aid alignment for actioning your insight-informed plan.

A report will do two things:

Some of the key components to include in your UX report are:

A key part of the UX research process is sharing insights with your team. It’s critical that stakeholders know the outcome of your research if you want to evangelize research and improve stakeholder buy-in.

Begin by laying out your research goals and demonstrating the connections to wider business objectives. Just like in your report, you should be speaking straight to stakeholders' pain points, hinting at how your user experience research will solve business problems. Tell them what you did and why you did it.

The next step of your presentation is to follow-up with key findings and atomic research nuggets. Created by Tomer Sharon and Daniel Pidcock, these digestible conclusions succinctly communicate exactly what you’ve done and the results you’ve reached while aligning them with product development. Sprinkling in atomic research nuggets also ensures that important results stay at the forefront of stakeholder’s minds.

Make sure to walk stakeholders through each step of your research process in a way that highlights how you got from start to finish. For example:

Consider creating a UX research case study with strong storytelling tactics and a memorable narrative. This provides ample context for stakeholders, making it easy to thoroughly digest what challenge you faced, how you dug deeper with UX research, and the insights you uncovered.

Hillary Omitogun, UX Research Consultant and Founder of HerSynergy Tribe, emphasizes how she personalizes reports based on the stakeholders she’s presenting to:

“I’ve noticed no one really reads long reports, so when sharing insights with the technical team (e.g. engineers) I focus on facts and hard data. While presenting to the marketing team, I include qualitative data.

For C-level executives and when I want a decision to be made, I go straight to the point in my report, adding direct & easily skimmable recommendations with hard data as support. For leaders or execs, I atie in recommendations to the business KPIs or goals.

Regardless of the stakeholder I’m sharing the insights with, I add an Appendix section with detailed insights.”

Conducting effective user interviews only gets you so far—you need buy-in to action your findings. A strong report is how you get them on board with your plans.

If you follow the steps above, you’ll be sure to transform a digital data pile into rich, descriptive insights, serving as the basis for actionable product development solutions.

To support you along on the way, we’ve spoken to UX leaders to collect their best practices for the analysis process. Here we go.

Confirmation bias is a cognitive bias that occurs when we have preconceived notions of what users think about a topic. Instead of objectively reading data, researchers are prone to projecting expectations on the responses, leading to an opinionated interpretation of results.

What does confirmation bias look like in action? Let’s say you personally worked for weeks on implementing a new dashboard for the user’s homepage, and you conducted interviews to check if users are finding the dashboard easy to navigate.

You might then be subconsciously inclined to read into ambiguous comments—playing up positive comments while disregarding negative experiences regarding the feature—due to your close involvement in its design and development.

To avoid cognitive biases in UX, the first step is being aware of them. If at all possible, ask someone from your UX research team to help as well—a second pair of eyes from a neutral party is invaluable.

Hillary shares her approach to mitigating bias when analyzing user interviews:

“With any type of bias, being aware of it is the first step. When we know such bias exists, we note it down and ensure that we remain open to being wrong.

“When I worked with team members on a project, I had everyone write down their hypotheses about the project, product, and/or users, as well as what informed said hypotheses or why they think so.

“This helps make sure we’re relying on the data and not what the Head of Product thinks should be done based on their own biases.”

Like anything, interview analysis is more efficient when you’re using the proper tool for the job. We’re not just talking about saving time and effort, either; while this is definitely one benefit, it’s only the tip of the iceberg.

Not only is manual analysis tough on time—it’s prone to human error. It’s easy to miss whole sentences while transcribing interviews, accidentally use one word over its synonym, or gloss over points of interests when assigning codes.

A strong user interview tool can help you avoid these issues. There are ample options to choose from, all boasting different features and functionalities. Some even support other UX research methods, such as running UX research surveys, conducting card sorting, and carrying out usability tests. The best tools do it all.

For example, Maze is a comprehensive user research platform that offers an Interview Studies solution. This user interviews research tool supports your user interview analysis with a stack of AI-powered features, including automated transcription, thematic analysis, interview summarization, and reporting.

Haley Stracher, Design Director and CEO of Iris Design Collaborative, shares her tool stack for running user research, including user interviews:

“I use a lot of systems like Hotjar and others that track user interactions post-launch. For user interviews during development, you can use products like Maze to not only conduct the interview, but also track the data afterward.”

Not all user segments will think the same thing. What may seem like an efficient and time-saving interface change to one user segment may appear confusing to another. It’s important to collect and consider demographically-diverse feedback from the get-go.

When analyzing certain codes, themes, and insights, always check the demographics and psychographic details of the users they came from. User personas are beneficial here, if you already have them. For example, you may find that it’s support and entry-level users that find a certain feature unnecessary while executives tasked with decision-making think it’s invaluable.

While user interviews can deliver a wealth of insights, they’re not always a one-fell-swoop solution to your UX research, design, and development needs. If you find your insights are lacking, schedule a follow-up session with willing participants. You can ask them follow-up questions to expand on certain answers for more context.

Hillary notes the importance of asking your users at the end of sessions if they’re open to help with follow-ups or future studies:

“Depending on the project or product, I often ask users if they’d like to be contacted for future studies. In a few cases, I’ve had to reach out to users to confirm that they said a certain thing if the recording wasn’t clear or I didn’t hear them properly.”

By this point, you’ve taken the insights from your research and presented user interview findings to stakeholders. They’ve listened to or read through your report and are fully on board.

Put simply, you’ve got the green light; it’s time to move forward with your next steps.

So, what is next?

User interviews are a starting point, not a finish line. The insights you uncover should inform further research, and act as a driving force behind your UX design and development initiatives.

You want to keep checking in with users throughout all stages of the product development process—from problem discovery, all the way through to post-launch. A UX researcher’s job is never really done—great UX research is a continuous process of speaking to users, listening, and helping iterate the product.