User interviews are a key aspect of qualitative research. There is such richness to be learned from listening to the words people use in a conversation and sharing an experience. A user interview creates a unique opportunity to  ask participants questions about their behaviors, thoughts, and feelings. It’s a timeless and invaluable way to gather data. 

This article examines how user interviews can be analyzed to create new knowledge using frameworks and methods that improve the outcomes of research. By the end, you’ll have several options in your toolkit for user interview analysis.

Why is it Important to Analyze User Interviews?

Taking the time to analyze user interviews is just as important as the decision to do user interviews in the first place. In a time or budget-constrained environment, you might think it’s at least better to do the user interviews, even if you don’t have the time to analyze them. However,  to avoid false signals, bias, or traps, it’s actually recommended that if you have to make concessions, talk to fewer people before you scrap the analysis phase. 

Doing user interviews and gathering data, such as notes or recordings, is only half the effort and really is more of a means to an end in the bigger picture. What you do with that data during analysis will make or break the outcomes of your research. For example, now that you have notes, recordings, and transcripts:

  • How do you break things down into smaller pieces of data, classify that data, and look for patterns? 
  • How can you poke and pull at your data to find tension or surprises? 
  • Which part of your data will you reference or call upon to prove your hypothesis?

It’s in the analysis phase that problems or the foundation of insights are illuminated. A methodical analysis process not only helps build trust with stakeholders, but drastically improves the quality of research insights. Trustworthy, high-quality analysis ensures that research helps discover new, compelling insights that reinforce the value and function of research.

Core Concepts to Consider During User Interview Analysis

There are a few concepts to consider when analyzing user interviews. 

  1. There are different types of user interviews to consider: structured and unstructured. A structured interview is a detailed approach that is more formal and intentional. It’s designed to gather a specific, evenly distributed dataset. In a structured user interview the goals are clearly defined, the user research questions are prepared ahead of time, and the interview environment and pace are intentionally structured to re-create a similar process and experience.

    An unstructured interview is more conversational. The interviewer might have a guide of topics to cover, but can ask open-ended follow-up questions, expand on important responses, and observers will often chime in with questions of their own. In this setting, the interview questions can be more fluid, meaning they might slightly change from one session to the next. An unstructured interview is a more casual option to consider when the goals are personal, the risks are low, or the budget or timeline is tight, meaning follow-up sessions likely won’t be a possibility. 

    There are many shades of structure and it’s often ordinal to the formality of the research topic. For example, if you were setting out to interview mentors about a topic for a personal learning or creative project, it could be advisable to improve your questions and process as you move through your group of participants, building upon the conversations you’ve already had. However, if you’re working on a software that detects brain pattern anomalies to monitor for seizures, you want to do your best to create a controlled environment with each participant, so that their responses and experiences can be analyzed against the others equally and in a pragmatic way. Goals, ethics, and expertise are all factors that should help determine the level of structure and rigor to bring to each type of user interview.
  1. The level of structure in the interview matters because it correlates with the level of rigor to bring to the analysis phase as well. When designing your analysis methods, choose a fidelity that corresponds to the research goals, the interview structure, and the overall environment of risk, safety, and importance. 
  1. One of the first things to come up as you move into user interview analysis is when to begin analysis. There is a growing consensus in UX research that waiting until your interviews or data collection is complete, before beginning analysis is a best practice. This helps deter early synthesis from swaying your interview process and creating a confirmation bias.  
  1. When beginning analysis, it’s helpful to reorient yourself to the problem. The goals of the research were likely stated in the weeks or even months ago that were leading up to kicking off research. Sometimes the overwhelming amount of data and emotional experience of the research process leads you on tangents and seeds of new information. While it might be helpful to note those tangents, for the research to be successful it should focus around the intended goals. This mental re-orientation might be as simple as re-reading the research plan to yourself or it might be a group discussion if you are doing collaborative research. This could also be a good time to “debrief” with the team and do a quick quality control evaluation to see if everyone is still aligned or if there are any reasons to tweak your plan for analysis, such as adding new tags to your project’s taxonomy or picking up an additional method. 

Create your own AI-powered templates for better, faster research synthesis. Discover new customer insights from data instantly.

How to Analyze User Interviews 

There are several guidelines that can help you analyze user interviews.

  1. Typically the data involved for user interview analysis is recordings, transcripts, and notes. Notably's interview transcription software allows you to automatically turn recordings into transcripts, but even if you have a note-taker or take verbatim research notes, the transcript and matching video or audio is an important artifact and data point to incorporate. Start by revisiting the interview transcripts to highlight the most important things that were said. Try and keep your highlights as atomic as possible, meaning 1 to 2 sentences max. A highlight or note should express a single idea, thought, or concept. 
  2. Use tools like Notably to classify, code, and organize notes that help you start to see the bigger picture. As you start to make connections and patterns emerge, you begin to get a good idea of the general thoughts and ideas that were shared.
  3. Use likeness to look for key concepts shared from multiple users. Identifying potential missed opportunities shared by many users can have big results.You can also use tension to inspire patterns as well. To do this, look for scenarios or quotes where a user says one thing, but does another. That divergence is often the intersection for new innovation or problem-solving.
  4. Also look for outlier data that could be relevant. For example, maybe not every user shared the same experience, but one user’s experience stands out for being particularly positive or negative. 
  5. Lastly, it's helpful to layer these patterns inside an additional framework to get an even deeper perspective. For example, now that you have quotes and themes, you might take a look to see how that data compares when looking at change over time or a chronological experience using a journey map. Or you might want to plot this data inside of an empathy map to see how those patterns drive the understanding of personas. There are many ways to get creative at this phase of the research and it’s often easier to do than you might think, especially because you already have a nicely curated dataset to work with.

Speed up Analysis with Notably

Notably is an all-in-one platform to help collect, analyze, and share research. All you need to get started is a single data file! You can add video, audio, transcripts, or documents right to your Notably workspace.

The fastest way to turn this data into information is with our AI-powered Summaries. You can choose a summary template that has been carefully crafted by research experts to tweak and modify, or use as is. From here, turn your information into knowledge by automatically highlighting and tagging your summaries, and push these notes to your Analysis board. (this is where the magic happens!)

The canvas in Analysis is a multi-dimensional workspace to play with your data spatially to find likeness and tension. Here, you may use a grounded theory approach to drag and drop notes into themes or patterns that emerge in your research. Utilizing the canvas tools such as shapes, lines, and images, allows researchers to build out frameworks such as journey maps, empathy maps, 2x2's, etc. to help synthesize their data.

Finally, One of the most powerful features in Analysis is the ability to generate insights with AI. Insights combine information, inspiration, and intuition to help bridge the gap between knowledge and wisdom.Even before you have any tags or themes, you may generate an AI Insight from your entire data set. You'll be able to choose one of our AI Insight templates to stimulate generative, and divergent thinking.

With just the click of a button, you'll get an insight that captures the essence and story of your research.You may experiment with a combination of tags, themes, and different templates or, create your own custom AI template. These insights are all evidence-based, and are centered on the needs of real people. You may package these insights up to present your research by embedding videos, quotes and using AI to generate unique cover image.

You can sign up to run an end to end research project for free and receive tips on how to make the most out of your data.

Want to chat about how Notably can help your team do better, faster research? Book some time here for a 1:1 demo with your whole team.

Give your research synthesis superpowers.

Try Teams for 7 days

Free for 1 project