How to do user research if you’re not a researcher

User research can be challenging and daunting, especially for small companies (like us). Here are some tips for how to run testing sessions on a small scale and still gain valuable insights from it.

A day full of user testing sessions is as exhausting as it looks here. (This is our lunch table by the way, the researcher and candidate are sitting in the room behind that wall.)

If you’re building digital products, you’ve probably heard that it’s all about continuous learning, validation, and iteration. In fact, you’ve probably heard it a million times. Put real product in front of real people. Test. React to it. This is how it’s done by the books.

In reality it’s not that easy. If user research is not ingrained in the culture of your company, it can be effortful and challenging.

We know that situation very well. A Color Bright is a small boutique studio for product design and development. We currently have a team of ten, and none of us has a business card that says “Research”. With that being said, we find user research to be incredibly useful every time we do it. It’s always quite the hassle but usually well worth it. This is why we wanted to share some experiences and recommendations with you.

1. Be clear about what you want to get out of it

Create a list of questions you want answered or hypotheses you want validated. You should not have more than three or four of them, and they should be as specific as possible. “Find out whether our app is any good or not” is not a great objective for a user testing session—the answer to that question will always be somewhere in between.

When compiling the questions and hypotheses, make sure you involve the entire team and all relevant stakeholders. Ideally, do this in a workshop format.

2. Recruit the right participants

You won’t be able to talk to an endless amount of people. Six to eight sessions across two days is about the most you will be able to handle and process. Therefore you need to select your participants carefully.

A premium candidate for a hula hoop research.

We usually start with putting together a screener form. Google Ventures has excellent resources on how to do that best.

Keep the form as simple as possible. You want to make it easy for people to fill it in (they are total strangers, why would they go the extra round for you?), and you want to be able to scan through the answers quickly. Eight to ten questions, that should be it.

Just make sure you get two things straight through the form: Are the participants really part of your target audience? And what would automatically disqualify them from the research? You don’t want to be sitting there after the session, bitching about how the person wasn’t qualified to judge your work.

As an example: We recently worked on a digital football publication featuring mostly heady longreads on stats and tactics. We wanted to make sure our participants were generally interested in this type of stuff, so we asked them to name the current clubs of three rather obscure managers. Nerdy enough? Great.

A good way to reach relevant participants is through promoted tweets. You pick a number of target accounts (“I want to test a football site, so I’m looking for people who also follow @Zonal_Marking and @sportingintel”) and promote a Tweet with a link to your research form (D’oh!). A $50 budget was usually enough to find at least ten excellent candidates within a couple of hours.

Incentivise accordingly to spark interest and avoid no-shows. People like cash (crazy, right?), but Amazon or iTunes vouchers also work and are easy to provide.

When scheduling your sessions make sure you factor in breaks generously. User research is exhausting and things are bound to go wrong. People are late. Prototypes break. Skype and Wi-Fi and microphones fail. And if everything goes right, participants will talk a lot and the session will run over.

3. Don’t let bias ru(i)n your session

If at all possible, hire someone who’s not emotionally involved with the project to conduct the session. This is to avoid leading questions, confirmation bias and general awkwardness. If you can’t hire anyone external, don’t pick the project’s lead designer.

Context is important. Explain to your participants who you are, what you want from them, and what you don’t want from them. But at some point you’ll have to shut up and let them do their thing. We’ve found that letting people interact with your app or prototype naturally and think out loud while doing so, is usually more valuable than asking questions. If you have specific tasks don’t ask how they would go about doing this or that, ask them to do it.

If you’re testing multiple variations or versions, make sure they are distinctly different from another and show them to different candidates in varying order.

Have the team (or representatives of the team) take notes throughout each session and highlight the most important bits straight away. Going through hours of recorded video footage is not something you want to spend your precious lifetime on.

4. Processing the session is a team sport

Compare notes as a team. The post mortem discussion amongst team members and stakeholders is the most important bit. One person alone can easily jump to premature conclusions (or wrong conclusions even). This is particularly dangerous if they’re personally involved with the project on an emotional level.

Discussion needs to be had, but not excessively. Talking to five people is not a scientific exercise per definition, so don’t obsess about finding the truth. Treat what people said and did as hints, not definitive answers.

Also don’t be bothered too much with formal documentation. There are enough PowerPoint presentations in this world, and the important bits will stick anyways.

5. Have one person in charge

As said before, user research is a lot of work. Scheduling appointments alone can be a headache, let alone writing real copy to replace the placeholders in your prototype, finding a quiet room, dealing with last-minute changes, etc. If you don’t want to halt your entire operation, assign one person to be in charge. They will need the rest of the team for sure. But they should drive the thing forward.

6. Be prepared for disappointment

Without a doubt the best depiction of what watching a research session feels like.

Some people will be frustratingly ignorant. Others will have reasonable criticism for you. But if you’re looking for someone to tell you how great your work is, you should ask your mother. Users failing to find what they are looking for, or finding the UI uncomfortable is exactly what you want from user research.

Now this sounds obvious. But if you have never done this before and are passionate about your work like we are, trust us: This stuff can be hard to digest sometimes. It really can feel like a slap in face. All these hours of thought spent on tiny details, and then these $%+&?# get it all wrong. But that’s how it is. That’s how the world is. And if you are willing to deal with it in the right way, research will not only be beneficial to the outcome but also motivating and inspiring as hell.

After all, there’s still nothing better than putting something out there as a team and watch real people enjoy what you built in real life.


Learn more about what we do over at acolorbright.com or follow us on Twitter.