Leading a set up of an efficient UX research infrastructure for the team.
Context
I'll start this case study with a note. Advocating for research is a nuanced undertaking. Researchers often say that UXR is treated as a waste of time. From my experience what drives this sentiment is an assumption that the best way to actually know-know is to Ship and Measure data.
That's where I'm hoping to flip the conversation: Ship and Measure is also a method of research and we should agree that not every change requires an extensive interviewing or testing as brilliantly demonstrated by Jeanette Fuccella.
Pragmatic and on-point approach to prioritisation of pre-release and post-release research was in the core of this project.
It's our ability to know what can be researched with limited number of people and what needs a critical mass of natural use that makes our research practice helpful.
Throughout this project my colleagues and I tried our best to establish a joint understanding of UXR across an entire organization. I believe we succeeded in some ways but needed more time to drive change in others. With this in mind, let's get to a more practical side of things.
What is UX research infrastructure?
UX research infrastructure empowers the team to research efficiently and in an organized fashion so that all research is accessible, sharable and reusable. As a member of the design team, I led the setup of this framework. I involved two other senior colleagues to make it into a collaborative effort.
In the span of several months, in parallel to other initiatives, we gathered and compiled best practices from the research world and our experience into an actionable framework. The final pipeline was heavily based on Lean UX for its pragmatism and iterative approach that suited our ways of working.
Goals
Our goal was to streamline research practices in the organization so that our team can make better decisions and de-risk assumptions early on.
Outcomes
More research projects took place since the release of framework and research made its way into product decisions. Time from planning to research execution decreased. Recruitment speed continued to be a major setback especially among enterprise users but it did improve comparing to pre-framework times.
Experience change
From
Disjointed research efforts, poor research stack and lack of process understanding led to low research count and insights being lost
To
Unified process and documentation for research with all essential guides and templates for the team
Trigger/Problem
We were facing quite mundane operational problems: research results and documentation were scattered around team's folders, there was no ready-to-be-used toolkit and user insights were collected in a random fashion making them biased.
In the meantime, our product teams often operated on assumptions and placed high bets without concepts and ideas being tested. There was a degree of research adversity among product people on both high and middle levels. Despite some resistance, there was an appetite to derisk and stress-test our assumptions early on.
Goals
Streamline UX research across the board
Upskill the team
Democratize research
Ultimately, help product teams make better decisions faster with the help of research
We wanted to the team to operate with a Lean methodology in mind so that high-priority research questions are put forward. As the infamous Lean UX canvas suggests, there are two questions that anyone involved in product should actively seek to answer:
What is the most important thing we need to learn first and what is the least amount of work to learn the important thing?
It doesn't quite roll of the tongue, does it? I feel you. As heavy as it sounds, it is still one of the most pragmatic intros to a research project that, as a bonus, resonates with a wider product crowd. We took it from here.
Solution
Lean became a motto of our framework. Another brilliant approach we used was Jeanette Fuccella’s model which guides the team through matching a challenge with appropriate design and research workflow.
How does a UX pipeline look like?
Process and decision tree
An outline of a process with all supporting documentation was created and shared with the team. The team also had an access to a Method Decision Tree so it’s easier to choose the right research method depending on the type of question, budget and timeline.
User insights channels
We mapped all the insights channels available to us: support tickets, past interviews and tests, surveys in tools like Hotjar and others.
Research stack
First, we collected requirements for research tools and once they were selected we defined the use for each research product we had. The ecosystem consisted of several platforms that had to be linked to each other. We created concise documentation for all tools and platforms with guidelines for the team. We used Heap, Dovetail, Metabase and Maze.
Research indexing and storing
It was important that qualitative research is indexed and processed in a way that any team member can access it and understand what was done, what personas participated and what insights were collected. To build a framework for continuous research research projects, interviews and insights should also be searchable – luckily tools like Dovetail do a great job with tagging user interviews.
In six months from now anyone should be able to find interviews and insights they are looking for, was the north star for insight indexing setup.
Templates
We provided the team with off-shelf that our team members have access to off-shelf solutions for various research needs: research plan, unmoderated and moderated testing, interviews, and more.
Method specs and guides
Each method had a profile with details and best practices to facilitate the process.
Recruitment
Finally, since recruitment can often be time-consuming we compiled guidelines on recruitment so this step doesn’t slow down the whole process.
Please feel free to inquire for more information about this initiative.
