You can’t do good design without doing user research. As UX professionals, it’s our job to make sure we produce good design. Ultimately, it’s the user who decides if we’ve done our jobs or not. Whether or not clients pay for user research, they are certainly paying for good design. So, here are my 10 tips on doing user research when there is no user research budget.
- Use existing analytics. If you are redesigning a site or product, review all existing data on traffic patterns, errors, and any survey results since the product’s last launch. Get a sense for what is going right and what is going wrong with the existing product. Don’t expect a single point of contact on the client side to have all the information you need. Do the leg work and pull the data together.
- Add the call center to your list of stakeholders to interview. If your client has a call center, interview the call center employees. Observe the call center at work if you can. Even just an hour is great. The call center knows better than anyone else in the business what the user problems and perceptions are. And when you interview stakeholders, ask them about analytics. Try to make questions open-ended so the answers can surprise you. Surprise is good.
- Identify key scenarios. Once you’ve reviewed existing analytics and aggregated stakeholder requirements, you can begin to sketch out key user scenarios. Make sure these scenarios come with measures of success so you can plan to gather analytics for the next product launch. UX is cyclical – line up your next success early. Ground scenarios in the return on investment they can demonstrate.
- Mock-up the concept by any means necessary. Sketches, HTML prototypes, iRise, Axure, comps; use whatever works for your timeline and your team to get the concept to a point that it can be shared. It’s better to get user feedback too early than too late.
- Users are humans. Remember – what you are testing for is human computer interaction. So, it is vital that you test with a human. Even if the user comes with a bias, even if the sample size isn’t statistically relevant, it’s better than nothing. There are humans around you. User research is always possible. If your client is targeting a particular generation, try to get users from that age group. But, first and foremost, test with a human. If you are going to use colleagues or friends as for testing, give them a persona so they have context for testing.
- Revise between tests. Don’t wait for multiple users to report a problem. If a problem is spotted in the first test, fix it before the second test. Adapt and evolve your concept rapidly and trust the process.
- Use the tools. Moderated, unmoderated, remote, field, lab; there are a lot of ways to structure a usability test. My personal favorite is watching users in their native environments. But there are a lot of tools out there to help you conduct remote unmoderated tests. They won’t give you much qualitative insight into user behaviors, but they basically run themselves and are way better than nothing.
- Don’t judge the user. I’m always surprised at what users have to say. I’ve seen users misunderstand scenarios and I could not fathom how they could be confused. That is why user research is so important. Take all feedback into consideration, even the seeming outliers. Stay objective. Don’t be afraid to get a little off topic. Sometimes the golden insight is found when the user drifts off your scenario. Trust the process.
- Don’t make it a separate deliverable. If your client isn’t budgeting for user research, don’t turn the research findings into a fancy deliverable. Take the notes you need and put your energy to use refining design.
- Defend design decisions. Armed with analytics, interviews, and user research, you will be able to objectively defend design decisions. It doesn’t mean you will win every argument. But you will be able to elevate your view to fact in a sea of opinion. Advocate for the user, always.
Are you a UX pro? How are you keeping users a part of your UX design process?
This is all good advice. I can’t tell you how many discussions I have been involved in where the UI/UX design is not based on proper research and analytics. Even when we would do user research it was such a small sampling that we needed to be careful not to over-rotate and make changes based on such a small sampling. The challenge we always had in UI/UX design is trying to figure out what design will work for most types of users…ie. the newbie, average and power users and also their job function within the product on another axis. Making things non-clutered but easy to discover features seems to be the trend that worked for us. Anyway, great article.
Interesting that you mention small sample sizes. I’m a fan of doing multiple rounds of qualitative user research with small sample sizes (3 to 5 users) and revisions in between. Even if I waited to test 15 users before revising once, it still wouldn’t be statistically significant. I’ve spent a lot of time discussing sample sizes with clients. As a colleague of mine said “If you trip on a rolling skate on the stairs, you don’t wait for someone else to fall before you move the skate.” With formative user research, I don’t think sample size should be the focus. Rather qualitative user research insights should be. You can get something more statistically relevant from analytics post-launch and line-up the next feature set based on those findings. All things are cyclical.