Atomic UX Research as a Research Framework

by | Jan 28, 2020 | General, Research

 | 6 min read

Conducting user research is the first step of every UX project. While exploring your users’ behavior, the data you collect piles up quickly. It is not an easy job diving into those amounts of data for finding a certain piece you need, answering your specific research question.

While working on our UX repository consider.ly, we’ve seen a fair amount of UX teams that came across this problem: A rich body of qualitative data accumulates but it’s not exploited in its fullest. It’s like looking through a telescope at a tiny fraction. There is so much more to explore right in front of you.

Now, what if you store UX knowledge in a way so that it is searchable and could be queried for your research question at hand? How can you build up a body of research that you can rely upon over time?

To cope with this challenge the concept of Atomic UX Research provides a solution.

What is Atomic UX Research?

Origin of Atomic UX Research

How to store UX knowledge in a meaningful way was once independently explored by two UX experts named Tomer Sharon and Daniel Pidcock. They came to similar conclusions and both framed the term “Atomic (UX) Research”.

Atomic Research is an approach to managing research knowledge that redefines the atomic unit of a research insight.

Tomer Sharon in â€śFoundations of atomic research”

Tomer Sharon designed and introduced 2017 a system called Polaris to follow the Atomic Research framework. He is former Head of User Experience at WeWork. Polaris is WeWork’s in-house platform to structure and retrieve qualitative research data.

Note that there is not the one definition for “Atomic UX Research”. As you compare the following with the groundwork of Tomer Sharon and Daniel Pidcock, you will see slight differences. We joined their models and adapted wording where suitable. For example, we view “facts” and “nuggets” synonymously, while you could also argue that “nuggets” rather refer to “insights”. We explain our reasons at the respective passages.

Definition of Atomic UX Research

The concept of Atomic UX Research is related to Atomic Design. In the Atomic Design, you can divide any UI into its single parts. This makes the reuse of e.g. buttons or checkboxes easier. On top of that, it maintains the consistency of the design as a whole.

So what if the principles of Atomic Designs are applied to qualitative UX research? Using modular research approaches can turn into a strong UX research framework. Atomic UX Research is a method for making UX knowledge management easier. It helps to organize what you know about your users by splitting information up into small units.

Dividing Content Into Nuggets/Facts

We call the smallest possible units of research insights “nuggets” or just “facts”. Users collect tiny bits of information. They consist of observations gained by research and their related evidence. There is no predefined format for the evidence. It can be user quotes, video or audio snippets, screenshots, photos, or anything similar.

“Facts” can be seen as a synonym for “nuggets”. There are two different terms due to the two original concepts. The Polaris framework shaped the term “nuggets”. In the end, both “facts” and “nuggets” describe the same atomic unit of information. We like to refer to them as “facts” since the word itself carries the concept of an immutable, non-debatable piece of data.

It’s All About Tagging

So what to do with all those facts? The key to putting single facts into a bigger frame is tagging. First, you set up the helpful tag categories for your project. Possible categories are:

  • Procedural (e.g. date, time, source, research method, evidence media type)
  • Demographic (e.g. age and location)
  • Experience-oriented (e.g. magnitude, frequency, emotions)
  • Business-oriented (e.g. revenue range, business unit, product line)
  • Service design-oriented (e.g. journey, act, scene, character, prop)

After that, you start matching your facts with tags from these categories. If you tag facts, you enable the retrieval of information.. In conclusion, tagging helps to combine and link facts from different sources. Thereby you set up the information structure for your research.

General Process of Atomic UX Research

For doing Atomic UX Research you should know the Atomic UX Research model. It is divided into four key parts: experiments, facts, insights, and opportunities. We explain each part in the following.

The Atomic UX Research framework Model
The Atomic UX Research Model
  • The first step, “experiments“, describes “what we did”. This means, that you start the process with studies that have been carried out. They could be user tests, surveys, and anything else that creates facts.
  • The facts you’ve gained from the experiments are unbiased. Because they are not something like ”the user said X” but a snippet of the user saying X. This step’s description is “what we found out”.
  • From facts, you get insights (“what we learned”). Insights are facts considered in context. That way you can interpret them easier.
  • The last part is identifying opportunities from what you’ve found (“how we could improve”). These new hypotheses can then be tested again to be verified or falsified as shown by the arrow in the graphic. Thus you should formulate them in a way that makes them testable.

A Practical Example

Now you know about facts, tagging, and the general process. What follows now is an example of how to apply this knowledge. Let’s imagine you are a user researcher. Your current project is to test a website for cooking recipes.

Graph of the Atomic UX Research in Practice
Atomic UX Research in Practice

Hence, your team conducts a survey and a usability test. When the experiments phase is finished everyone starts evaluating the results. All researchers start looking for relevant facts and tag them. Some facts you find out about the website are:

  1. 1 user accidentally clicked the “Add to favorites” button
  2. 2/4 users did not find the “Add to favorites” button
  3. 1 user says she would like the possibility of adding personal notes to recipes
  4. 1 user thought the comment section would be a section for personal notes

The next step is to derive insights. Therefore you put the facts into context by combining them. In sum, you have findings regarding the “Add to favorites” button, the functionality of adding personal notes, and the comment section. Those insights make you think that:

  • The “Add to favorites” button is badly placed
  • A function for adding personal notes to recipes is missing
  • The comment section is not clearly marked

From these results you identify the following opportunities:

  • Improve placement of the “Add to favorites” button
  • Add a section for personal notes
  • Improve indication of the comment section

After that, your team can test each of the opportunities. This means restarting the whole process and carrying out experiments again. In doing so you can improve your cooking recipe website step by step.

Which Tools Support Atomic UX Research?

The following section will describe two ways of how Atomic UX Research can be done in practice.

Atomic UX Research with Spreadsheets

A simple way to carry out Atomic UX Research is by using spreadsheets. Thereby a table serves as a database for facts. In the first column, you write down the facts. In the second one, you fill in the insights. The third and last column contains opportunities.

Within this spreadsheet, your team collaborates. Each member documents their findings and tags them. When customizing your spreadsheet take care of two tagging rules:

  1. The same tag must be usable for several facts.
  2. An individual fact must be taggable with more than one tag.

The Design System team could then have access to the Atomic UX Research spreadsheet and assess what needs to be improved. Doing Atomic UX Research with spreadsheets is easy and fast to implement. Once set-up, you can reuse this spreadsheet for different studies or rely on templates like this Airtable example for the Polaris framework. Nevertheless, amongst other helpful features, a visually appealing overview is missing.

Atomic UX Research with a UX Research Tool

Another more powerful possibility is using special software that was built to support UX research systems. These UX research repositories offer a wide range of functions such as:

  • Visual overview of your facts and insights
  • Central repository for different kinds of data (texts, pictures, audios, videos, etc.)
  • Dynamic repository with functions for linking, searching, and filtering
  • Working directly with the collected material (like video snippets) instead of quoting users
  • Consistent use of tag taxonomies across your team
  • Features for supporting the analysis, such as interview transcripts or sentiment analysis

10 Reasons for Using the Atomic UX Research Approach

To end with here are ten advantages you benefit from when using Atomic UX Research methods.

Atomic UX Research …

  1. structures collective knowledge
  2. makes UX knowledge visible
  3. provides dynamic storage for your findings
  4. easily lets you include stakeholders
  5. accelerates workflows and increases efficiency (especially when working with distributed UX teams)
  6. improves searchability by browsing for tags instead of mining through data
  7. prevents redundancy
  8. makes verifying your findings easier
  9. removes personal bias
  10. allows saving information for future use

All in all, Atomic UX Research allows considering the UX research processes as more coherent than before. With this concept, your research and search for insights are no longer limited to your current project. And when you think about your research’s long-term effect: The more research you carry out, the more facts you combine, the more holistic your UX knowledge will become.


Learn more about how Atomic UX Research works and how to apply this framework in consider.ly:

consider.ly is a fast-growing tool for quali­tative data analysis and UX research repository.

Follow consider.ly

Isabelle Woytal

Isabelle has experience in different kinds of usability testing and thinks putting effort in UX makes the world a little bit better.
Share This