Discovery research: What it is (and isn’t) and why it matters

Discovery research (also called exploratory research or generative research) involves investigating the problem space in order to gain a deep understanding of the people who might be using a product or service. help teams design products and services that actually help people do things. 

To borrow a definition from NN/g: 

A discovery is a preliminary phase in the UX-design process that involves researching the problem space, framing the problem(s) to be solved, and gathering enough evidence and initial direction on what to do next. Discoveries do not involve testing hypotheses or solutions.

Through user interviews and observational, ethnographic-style research, discovery research reveals helpful insights about specific problems, how users deal with them, and additional details about users’ lives, desires, motivations, and more.

This process helps teams uncover the best solutions, using “how might we” statements to brainstorm ideas. 

Generative research ≠ evaluative research

It’s important to make a clear distinction between what is actually discovery research, and what might be masquerading as discovery research. Sometimes, what seems like discovery research is actually evaluative research—like requirements gathering or the iterative research and testing that goes into optimizing an existing solution.

When used in the discovery phase, evaluative  research methods ignore the full array of possibilities, and focus only on what’s already in front of you, and what you *think* you already know. This research tends to be geared more toward validating decisions you’ve already made, rather than generating new solutions. 

The value of true discovery research—which should be ongoing throughout the product development process as products and audiences evolve—is its ability to save you from the danger of making assumptions based on anecdotal evidence or the CEO’s personal opinion. 

Using discovery research to unearth evidence teaches you about what you actually know and what you don’t know. This makes it much easier for your organization to identify and develop solutions that work. It also helps clarify what’s at risk (wasted time and money, maintenance concerns, feature bloat, etc.) if you move forward with a project and it fails. 

Critical steps for getting started with discovery research

1. Get stakeholder buy-in

One of the common roadblocks holding teams back from doing discovery research is a lack of buy-in about the importance of exploration. According to Maria:

“A lot of people I speak to really want to do discovery research. But they can’t because other people in their companies don’t understand the benefits of discovery research. These people think they already know everything about their users and problems and which solutions will solve them.”

To clear this hurdle, UX professionals need to invest time and effort in two areas: evangelizing discovery research and asking the hard questions that expose the need for it in the first place.

Asking hard questions might be uncomfortable at first, but it often turns out to be a great starting point for valuable conversations. Maria recommends speaking up in the moment.

“UX practitioners, designers, and other people working in product teams need to point at things and ask, how do we know this, where is this from, where is the evidence that says this is the right solution?” 

It’s also important to address the big-picture issues. “The real question is: how much risk are you willing to carry forward?” Maria says, referring to the role discovery research plays in mitigating the risks of uninformed decisions. The larger the project, the more time you’ll want to spend in discovery because the exposure is that much greater. 

When it comes to evangelizing discovery research to other team members and managers, Maria recommends a collaborative approach. 

“User research is a team sport. It shouldn’t be the case that you are going away and doing all the research and then just reporting it back to people. That doesn’t work well. It’s much better if the team comes along and observes the research first hand.”

To that end, UX professionals should invite people into the process, so that they can experience it (and see the benefits) first hand. This helps non-UX folks internalize the knowledge and increase their empathy for users. A pragmatic approach that starts small but delivers ongoing insights  is a great way to engage others early in the process, and guide them through the entire journey. 

2. Set clear goals

Starting a discovery research practice doesn’t need to be a huge undertaking. Discovery research can easily be adapted to your particular context and rolled out incrementally. You can start with light-touch activities, and then scale up as needed. 

For instance, if you already have a strong knowledge base about your users, you can focus on simple and manageable activities that build on what you already know to expose new opportunities and risks. 

The first step in any generative research project is to set concrete goals. You need to reverse engineer these goals from the ultimate objective: To provide the team with the ability to make a decision. 

“It’s important to specify the decision you need to make. And then the next step is defining which pieces of data you need to inform that decision—what questions do you need to ask, what answers will put you in the position to make the decision?” 

Being really clear about objectives is crucial because measuring progress against your stated goals is the only way you’ll know you’ve done enough research. Otherwise, you could go on exploring the problem space indefinitely. 

“Unfortunately, there isn’t really a golden number with qualitative research. You just want to ensure that you’re recruiting a representative sample, and that depends on the context of your project and the diversity of your use cases and users.” 

The good news is that discovery research doesn’t require a huge sample set. You can look at smaller groups, and still successfully identify themes and trends. Sometimes, you only need to talk with a few people to get a clear signal.

“If you’re continuing to do interviews, but have exhausted all the possible personas and are not learning anything new, you can probably stop. If you haven’t recruited a representative sample, and are seeing spotty results, you need to talk to more people.”

3. Choose your research methods

There are two broad categories of discovery research methods: interviews and ethnographic studies. For the full picture, Maria recommends combining them:  

“Each method provides a slightly different piece of the picture, and they complement each other really well. And because we’re looking at really small samples, you’ll feel a lot more confident when you see similar themes emerging across different research methods.”

User interviews are incredibly useful for capturing attitudinal data that reflects subjective experiences. This is particularly valuable if you’re building customer journey maps, service blueprints, or experience maps. Interviews can tell you what people go through to achieve a specific objective, the challenges they face, and how they deal with those challenges. 

Ethnographic research, on the other hand, is observational. It can be as simple as observing users in their own environment and capturing specific patterns and behaviors (via notes, photographs, video, etc.). It might also involve semi-structured interviews. Because this type of in-person observation can be difficult during a pandemic, you might choose to employ digital ethnography techniques using remote tools or have research recruits participate in a  diary study. 

While some folks might also include focus groups and surveys in the  discovery research phase, these methods can be hard to get right. Maria is not a big fan of either tactic. 

🔌 Recruit, manage, and stay in touch with research participants

User Interviews makes it simple, whether you're conducting generative interviews or running ethnographic studies. Plus, your first three participants are free!

The time for discovery research is now

Twenty years ago, emerging UX professionals were fighting for buy-in on usability testing. Today, user experience designers and researchers—and the organizations that support them—realize that UX needs a seat at the table a whole lot earlier in the process. As Maria puts it:

“We’ve gotten good at building the thing right. But we’re not necessarily as good at selecting the right thing to build.”

Now is the time to help organizations understand the discovery advantage. 

“On paper, it looks like we’re asking stakeholders and product teams to spend more time on an activity that isn’t delivering anything.We need to do a better job of communicating the very real advantages of discovery research. We need to help them understand that this is an important step toward reducing risk while allowing us to move forward with greater confidence.” 

Listen to the full episode (with transcript):