Case study: Advancing UX Research at Elli

Introducing mixed-method approach and changing perspectives

Myroslava Domanitska
The Startup
4 min readDec 19, 2020

--

Elli, a brand of Volkswagen, is a company that provides energy and charging solutions to consumers and businesses. One of the products Elli offers, and the one I worked with mostly, is Elli app which allows electric car drivers to charge their cars at public stations or at home. The app is currently available in Germany on App Store and Play Market.

On my 8-month journey as UX Research Intern at Elli, I went through a researcher’s ups and downs — but also learned a great deal about what works in user research (and what doesn’t 🙂).

During my internship, I was lucky to collaborate with incredibly talented designers, developers and product managers. Together we worked on making Elli e-mobility solutions even more user-centric, developing new and improving existing products, creating a shared understanding of who our users are and what they really need.

Research process as-was

Methods used: interviews, moderated usability tests

  1. Initiative to conduct research generally came from designers and was motivated by a sincere curiosity about what users think and a need to validate designs
  2. Research was designed to test users’ specific touchpoints with the product, those that we felt might be confusing or frustrating
  3. Each research project took us at least 4 weeks, including preparation phase, recruitment and analysis
Steps in our research process as-was
Research process as-was: We went straight into deciding on research methods and conducting research.

What we introduced

1. User research as a shared team activity

We work closer with product managers and product owners at the early stages of product design to integrate user requirements early in the development process and to motivate product teams to perform user research more often and iteratively. Everyone at Elli, regardless of role and responsibilities, is encouraged to plan and conduct user research.

Now we always set assumptions before planning research and get back to them after analyzing the findings to learn. For example, we found how important it is to write down what we think users are going to tell us before we start interviews: what could be their goals? pain points? needs? This also helps involve the entire team in research planning early on.

2. Defined which methods we want to use for which purposes

We realized the three key goals which Elli seeks to fulfil with user research:

  • to understand who early adopters of our solutions are, to learn about their motivations and needs (e.g. early discovery research for new products). Methods: interviews, observation, contextual enquiry
  • to test new flows or design explorations (e.g. at the early stage of digital product design or when developing new features). Methods: moderated testing
  • to quickly test updated designs before handing them over to development (e.g. when re-designing existing flows). Methods: unmoderated testing, A/B testing

Later I found out that Nielsen Norman Group classifies research methods in the same way — so we went on to see which methods are missing in Elli UX research. The first two goals we generally covered with the methods we used. We introduced A/B testing to quickly test designs and inform our design decisions with quantitative data — here you can read a case study about how it worked.

3. Visualizing user research findings

We turned Elli’s internal Design Wiki page into a discoverable user research repository to make sure employees always have a chance to read through some exciting insights presented in a playful form. We present text- and quote-based reports visually, e.g. as affinity maps or exciting user stories.

As user research advanced at Elli and became a more natural ongoing activity, we felt like there was a need to link all the findings. We wanted to make it easier for teams to incorporate those findings in their design process as well as build upon existing research — that is how we identified the need of a full-scale user research repository, which Elli plans to set up soon.

Advanced research process

Methods used: a combination of qualitative (e.g. interviews, usability tests) and quantitative methods (e.g. A/B testing)

  1. User research is an organic process that is by default planned at the required stages of product design. It involves not only designers but entire product teams, which also helps us set more specific goals and expectations
  2. We developed a more comprehensive approach to user research, where we regularly involve users to test explorations and ideas before we invest time in developing them
  3. We introduced new methods that would naturally speed up the research process since there is no need for moderation and a faster recruiting process. That would allow us to quickly get insights and iterate faster
Steps in our advanced research process
Advanced research process: We introduced goal and hypothesis discussion before defining research methods, as well as checking hypotheses before presenting results.

What I learned

  1. At human-centered organizations, the need and motivation for user research come naturally. At Elli, I was lucky to work with colleagues who saw real value in every story users told us. That’s a prerequisite for setting up a research process that works well — a case in which a little tuning of our approach can create magic ✨
  2. In such a dynamic company as Elli, there were many initiatives and other tasks that deserved attention — that showed me how important it often is to internally sell user research projects, both to encourage employees to participate and support (for future projects) and to attract attention to findings (for past projects)

--

--