October 23, 2022

How to use tags to better analyze user interviews

Tags help companies categorise and extend the lifespan of insights gleaned from customer conversations. In this post we explore different tactics to implement a robust tagging taxonomy across user interviews and other customer conversations, whether you are a UX Researcher, founder, or other customer-facing professional.

UX Research

Nowadays, professionals that often talk to customers as part of their job (e.g. UX Research, founder, Customer Success) can do so virtually and at scale using popular tools like Zoom or Google Meet.

However, these individuals and their companies are now facing huge volumes of raw data coming from these video calls. How should they uncover, save, and share meaningful insights in this sea of noisy data?

The answer to this question is to use tags.

Tags enable customer-centric organizations to quickly and easily access key insights from their user interviews and other customer conversations. This gives them a better understanding of their customer’s pain points and aspirations.

Moreover, tags that stem from a coherent taxonomy help companies answer critical research questions. Being able to answer these questions help companies become more customer-centric and deliver the right solutions for their customers .

In this article we will cover the following points related to tagging:

  • What is a taxonomy and why do we need one ?
  • How to build a taxonomy ?
  • How should you use a taxonomy ?

Introduction to Taxonomy

A taxonomy is a way to classify information that falls under a given topic. This enables companies to group information that is related under the same category. Therefore, a taxonomy represents the vocabulary used by customer-centric organizations to name and organize crucial information.

A taxonomy encompasses multiple tags that are related. While a taxonomy represents a higher level of abstraction, tags on the other hand are the manifestation of a grouping when applied to an observation.

These elements are core components of what we refer to as a nugget of insight. This term originates from the UX Research world and was coined by Tomer Sharon, who also popularized the concept of Atomic Research.

In Atomic Research, a nugget is the smallest unit of insight from qualitative data that customer-centric organisations can leverage. A nugget is composed of:

  • an observation: this is a statement made by a participant that illustrates a problem, validates an idea, or introduces unexpected and important information
  • the evidence of that observation: when it comes to video calls this is a short clip from the original video that provides a proof of the observation
  • a tag to categorise the  observation

For instance, imagine you conducted 3 usability tests over video and have identified these 3 excerpts from their respective transcripts:

  1. ….I am trying to find the next button, but I don’t see it…
  2. …. Oh! I forgot to change the address, how do I go back?….
  3. … I feel lost! What do I tap to go to the homepage…
Example of how to create highlights with tags in EnVsion

What do all these examples have in common? Answer: they are about usability issues.

These 3 excerpts could thus constitute 3 nuggets from the same insight with the tag usability issue. Tagging insights like this makes important information around usability issues easy to organise and of course retrieve for anyone with access to the system where this information is stored. These systems have traditionally been called repositories.

Benefits of using a taxonomy

Adopting a well-defined taxonomy together with clear tags yields the following benefits for companies:

Avoiding knowledge silos

  • Data is not scattered throughout the organization and is instead grouped around well known categories
  • With a taxonomy, anyone in a team can access relevant data whenever they need it from different participants and sources.

Increased life expectancy of customer insights

  • The insights collected over time could be applicable across projects for years to come.
  • Having a good taxonomy helps you make use of existing knowledge by filtering out insights from various studies.

Stakeholder buy-in

  • You are more likely to influence stakeholder decisions since you can provide compelling evidence of your claims around a particular topic

Related information is easier to retrieve

  • Similar to the way books are organised in a library, a taxonomy provides the index for you to easily retrieve nuggets from similar insights
  • Gone are the days when you had to copy and paste Google drive links, and don’t need to repeat the research.

What are the different types of tags?

To build an effective taxonomy, customer-centric companies must be aware of the different ways tags can be created to determine what approach works best for them. Firstly, tags can either be global, which means they could be applied to any material, or project specific. Additionally, tags can be derived either deductively or inductively.

Global tags

Global tags are used by everyone within the organisation and across projects. The taxonomy around global tags is developed after interviewing multiple stakeholders across different functions within a company to understand what topics they would like to capture insights for.

Examples of global tags could be pain points, usability issues, motivation, etc.

Project specific tags

Project specific tags are related to a single project. When working on a given project, there is a lot of information that is unknown ahead of time. Project specific tags therefore are useful  because they help identify themes from project-specific conversations.

These tags can be created with more flexibility that global ones. For instance you could define these tags without seeking consensus for the time being. It’s important though that in planning these tags you make sure that some of them are related to the goals of the project. You should also leave some room for tags that highlight unexpected insights.

Examples: applications used to book flights, reason for buying milkshake in the morning, etc.

Inductive tags

Inductive tags are created on the go, without advance preparation. They reflect observations they could not have been predicted and planned for ahead of the conversation. For instance, they give a UX Researcher analyzing a user interview room for manoeuvre so that they can capture surprising yet important information outside the initially defined scope of a research project.

Deductive tags

Deductive tags refer to any tags that are created prior to a conversation. Whether these tags are global or project specific, the taxonomy for these is agreed upon with team members and stakeholders to ensure they use a common vocabulary. This common vocabulary provides a consistent way of grouping and retrieving related insights.

How do I make my taxonomy?

We recommend you follow these to get started with defining your taxonomy:

Step 1 : Define global tags deductively

There are multiple ways in which you can define global tags in advance.

Tags derived from Stakeholder Objectives (Questions)

Objectives are things that your stakeholders plan to achieve finding in your repository.

Simply start by talking with your stakeholders to try understand what they would want to get out of the repository. We recommend you conduct private interviews with each stakeholder individually.

A product manager might ask, what features are people requesting? what bugs were they reporting? What suggestions do people have, etc.? You can then extract tag names out of it. Feature requests, suggestions, bugs, etc.

Tags derived from Use cases

Retrieving information can also be based on use cases, workflows, or the journeys a user might make within the app. Accessing insights in such a fashion of course is highly dependent on the type of organization and the project that you are working on.

For instance, in the case of a food delivery app, we could have the following tags based on what stakeholders are trying to find out regarding use cases: ordering due to of lack time, ordering to unwind, ordering as habit, ordering for romantic setup, etc.

Tags derived from Interview questions

These are interview questions derived from an interview guide.

For instance for the question  “What are the reasons for installing xyz software ?” you could create the tag reasons for installing xyz software.

In general, it’s a good idea to have tags that are expressive so that it is easy to understand what the content tagged is about without overthinking.

Step 2 : Test with existing data

Now that you have a set of predefined tags, it’s time to put it to test.

  1. Collect a bunch of sample data from different sources, and see if they fall into the tags that you’ve created.
  2. This will help you find tags that are relevant and those that are not.
  3. If you find anything new popping out, there’s always scope to define more tags.
  4. Now assess with your team if these categories make sense to you!

Step 3 : Create a dictionary

A dictionary is a place where every tag you create is defined. Tags fall into many groups, some generic and some specific to your project. Granted most of your tags should already be in sync with the vocabulary that is used in your organization, but a dictionary acts as a place for anyone to use as a reference.

  • Get all your stakeholders involved and collectively build a dictionary
  • Any new project specific tags can always be added
  • A dictionary is ever-growing and evolving

How do I know what to tag?

You shouldn't be tagging everything. The most important things to tag are

  • Quotes around values, beliefs, opinions, and attitudes
  • Anything that the user thinks out loud while performing the task
  • Emotions and sentiments that you observe
  • Anything related to the user’s vocabulary
  • A new or surprising observation that moves your research forward


Being able to retrieve valuable insights and acting upon it is what turns businesses into customer-centric organisations that deliver what the customer needs. But in order to have a pulse on these insights it is critical for companies to have a consistent way of grouping related observations.

That’s exactly what tags are for. Customer insights repositories like EnVsion enable you to group every relevant observation using tags and derive insights from these. Tags extend the life-expectancy of these insights, make information quicker to access, and by extension help create ever green company knowledge.

There isn’t a singular way to define tags, but we recommend you speak with stakeholders and colleagues to understand what they want to get out of your conversations. Once this preliminary internal research is finished you should be able to define the original tags for your research.

Remember that your first iteration won’t be perfect. Continuously seek feedback, and keep refining your taxonomy. You are well on your way to catch all the golden nuggets that will help your company ship delightful experiences for your customers.

Read more
You might also be interested in these
New Feature
Save time analyzing user interviews with our Highlights Wall feature

Eddie Forson

We've just released a new feature on EnVsion to enable UX Teams to save time analyzing and synthesizing insights from their user interviews and other customer conversations. Read more to learn about our Highlights Wall feature.
UX Research
Why you should start with generative research

Eddie Forson

This post covers how UX teams in customer-centric organization should approach generative research in order to uncover real user needs and develop compelling solutions that address those.