One of the guiding principles for the team at Repositive is to build something that solves a problem and is actually useful to people. Therefore, we are committed to working with user centred design (UCD).
"Usability goals, user characteristics, environment, tasks and workflow of a product, service or process are given extensive attention at each stage of the design process."
In this post I wanted to explain the trials and tribulations we have faced in trying to get user insights over the last couple of years, and the user research strategy we have adopted to get frequent user feedback.
A Niche Target Audience
We are building a product for a niche and small target audience. We estimate there are 500K – 1M researchers in the genomics field, this means that, though we intend to expand in the future, with our current offering we probably have a potential user base of 100K – 500K. These people are scattered across the globe, in different institutions and companies. Having niche target audience also makes it difficult to find users for usability testing.
A user persona is a representation of the goals and behaviour of a group of users. We synthesised these personas from data collected from interviews with our users. Getting into the mindset of our personas can help guide our design and product decisions.
The problem - How it was in the beginning
We want to work with UCD and have user input, but… it's hard!
When I first started working at Repositive two years ago, our user research and UX sessions were infrequent and sporadic (about once every 2 weeks). They were mostly performing face to face interviews which lasted about 1 hour. The infrequent UX session wasn't a huge issue as we weren't building very fast at that point, but what we did need, and lacked, was lots of user insight. We couldn't get users' working environment and understand their current workflow which meant user feedback and insight was not integrated in the product development.
The Solution - The User Council
What is a User Council?
Taylor Wescoatt, one of our mentors from Seedcamp, suggested we try out a 'customer council'. A customer council is a group of people who are actual or potential consumers of your product who are willing to give you advice in your early stages. Typically, it’s because they are early adopters and love what you do, and will do it out of the goodness of their heart (think open-source). This is qualitative feedback, and even though it is somewhat biased it will be valuable to you.
The customer council is a good idea as it reduces the stress of recruiting new users every time you want to test something. It can also help you build a passionate user base, that results in:
- A source of valuable information.
- A network that can eventually drive revenue.
As the Repositive Core platform is free, we don't have customers, instead we have users. So we decided to call it the User Council.
With the user council, we get consistent and frequent user feedback, which means we can start to iterate and develop fast.
In total we have had three user councils so far, we set up our first user council May 2016. The user councils typically contain 10 – 15 people. These people are committed to give feedback for a 6 month period. Over this time we preform 1 contextual interview and at least 2 ‘sessions’ per month (more on this later).
The first time we built a user council it was really hard to recruit people as we had such a small user base and had no network. We contacted people from all channels asking them if they would be interested:
- People we had met at conferences.
- Those who had done UX sessions with us.
- Researchers who had given us feedback.
- From our current user base.
One of the main challenges was to figure out what incentives we could offer. We had no money and no product, so we couldn't pay them or give them free subscriptions. Instead we offered the following:
- The opportunity to be part of a growing startup.
- For their voices and input to be the foundations of the product we build.
- We would personally help them find the data the needed for their research.
- To get awesome stash.
The main form of communication, which is almost 100% effective is email. It seems most researchers use email, a lot, and respond fast.
Each time we set up a user council we also set up a slack channel. This gives the members a chance to contact me but also to chat among themselves. But something I've learnt is that the majority of our users don't use Slack day-to-day and don't use it to chat among themselves. A potentially interesting finding in itself; researchers don't like instant chat apps?!
We use spreadsheets to stay on top on things, to know who are the members of which council, when we last chatted with them, and what we have received feedback on. These are in the cloud and shared within the product (me) & design (Jana) teams so we can stay up to date.
What we use it for
Contextual interviews. These are about 1 hour long where we interview the researchers about their workflows and how they interact with genomic data. This is ideally performed in their workplace so we can also understand their environment and distractions. These interviews help us to generate ideas and see how we can build new features.
Surveys. When we have specific questions and we need answers fast we just send out a quick survey. We keep these very short and simple, this means people usually responded very fast as they have learnt that the surveys are quick and easy.
Usability testing. Once we have concepts, designs and prototypes we do online UX sessions with the council members to see how they interact with them and to get their feedback.
Evangelising. Our user council members spread the word, they talk about Repositive at events and with their peers. We send 'stash packs' to the council members so they can hand it out to their colleagues.
“The last half a year went very fast, indeed. I would be interested to stay in the user council :) It is an experience for me personally as well”
5 great things about a User Council
They are reliable. If we need input or feedback we can just call on them and we know we will get an answer.
We build relationships with them. This means they continue to help after they are no longer in the council and they become part of the company's fabric and history.
By getting to know them we get a much richer idea of their background and the context of their feedback. This means we can infer much more from the behaviour of our other users.
We work with them over time so we start to see how their workflows change over time. Sometimes they spend months applying for grants or writing up publications and wouldn't be searching for data. This helps us understand traction and churn within our user base - just because someone doesn't come on platform for a few months doesn't mean they have churned, they'll come back.
- We get bonus feedback. Our user council members will get in touch even we aren't asking for feedback. If there is a bug, or they have ideas for the product, or there are events they think we should attend.
"I have another suggestion, you can consider it based on your priorities" - Roy
"I just have quick idea about the dataset arrangement" - Mahantesh
It's not all plan sailing though, there are a few things to keep in mind when setting up your own council.
The time it takes to set up the councils initially is not insignificant. It's a lot of work, even if you aren't having to hunt high and low for interested parties. You need to vet potential members, send emails and have calls to explain the concept and onboard them.
It will be almost impossible to avoid the positive bias. The people who are signing up to be part of your council are interested in what you do, they care about your vision and mission. Therefore, there will always be a positive bias, and you have to keep this in mind.
As the council is only a limited number of people there is a lack of breadth. You are only getting feedback from a handful of people. This is another reason to know who your target audience is and focus on those people on your council. Input from other 'potential future' users may seem like a good idea, but it will dilute your results and you will struggle to focus on what to build.
Finally, it's not scalable. The time, and manual input that is required means you can never maintain this relationship and involvement with more than 10 - 15 users.
Being involved in a growing start up is a big incentive.
Don’t do all the contextual interviews at one time. They take a lot of time, if you have other work to do it will become unmanageable, spread them out.
Short, remote UX sessions over skype work.
Building a relationship with our user council members gave us a much better understanding of their workflows than we ever got from user research interviews.
- It can be easy (and fun) to integrate user's feedback into the development of a product :-)