Image by Mark Stuckey

Interview with Dovetail’s CEO, Benjamin Humphrey, about making the most of user research.

How would you describe an ideal user research process? What are the key research methods, and when should they be used in product design?

I’m a big fan of Erika Hall’s Just Enough Research and Continuous Discovery Habits by Teresa Torres. I like interviewing customers to figure out what’s going on. There’s this great quote from Jeff Bezos where he says “when the anecdotes and the data disagree, the anecdotes are usually right.” I tend to lean more on qualitative research than quantitative data.

Co-design sessions are also underrated. It’s important to bring customers to the office and actually do workshops with them, such as collaborating when designing a new feature. Customers can talk about the problem they experience, and designers can sketch something right at that moment on a piece of paper or whiteboard and ask customers, “Do you mean that?” The customers will be able to evaluate the solution in real-time. We did a lot of co-design in the early days of Dovetail and still do.

What common problems related to research do you notice in many organizations?

One problem I see often is that sales and marketing teams in many organizations are disconnected from how the product is built. Useful insights that the sales and marketing teams collect are often unavailable to the design and development team, and vice versa. Dovetail aims to solve this problem by creating an insights hub to store sales conversations, customer success calls, market research, and more. This database allows people from different departments to access this knowledge and learn about users and their needs.

What are your suggestions for facilitating collaboration when you invite people from different teams to a co-creation session?

Driving collaboration is tricky. It is broader than just research. A combination of rituals and tooling can help. Rituals might be more interesting. The practice we follow at Dovetail is every Friday morning, we do a company-wide demo. For an hour, people share what they have been doing during the week. This is a totally voluntary activity. Our research team and customer success team use it as an opportunity to share valuable insights about our users (for example, a call recording that sheds light on an important topic). By bringing customers in front of people, it becomes much easier to communicate important ideas. Since we’re doing it on a regular basis, it helps to keep everyone in our organization informed and aligned.

Our research advocate also runs a monthly customer showcase for the entire company, where they share latest insights in Dovetail explaining specific topics. The topic can be “how to satisfy enterprise customers” or “how researchers can use AI in their workflow.”

Last but not least, I believe that blogging can also help facilitate collaboration. The great thing about blogging is that you can frame a specific problem as a story, which can encourage the reader to learn about the topic. Good storytelling also helps to build an emotional connection with the user.

Do you have any suggestions on how to avoid outdated information in your insights hub?

It’s important to assign the proper metadata to your research so you can, for example, filter the data in your knowledge base by date range (which Dovetail enables). However, historical data can also be valuable. For example, you can watch a video on YouTube that is ten years old and still get valuable information from it, right? While some user insights from ten years ago might be irrelevant (if users mentioned usability issues in your product ten years ago, there is a high chance you already solved them), strategic learnings can remain relevant for many years.

Suppose you have a large research base with insights you’ve collected over time. Do you have a universal approach to how team members should interact with it to extract the most important insights they need now?

First, an insights hub should not be a collection of dusty reports created over time. It should be a database of insights about user behavior. Team members should be able to quickly filter all available data in this database to find what they’re looking for. But the task of filtering should be intelligent, and this is where AI really shines. In Dovetail, we recently released a new feature called magic search. You can write a prompt (a question you want to find an answer to), and the system will analyze all available data in the knowledge hub and offer you the answer. The insight will be created at the moment of consumption (when the team member or stakeholder asks the question).

Do you have any recommendations on how to present research findings?

I like the inside-out approach, where you have a knowledge base that collects all signals about users and acts as a data warehouse, and you extract relevant insights exactly when you need them. Think Snowflake or Databricks for qualitative insights. Instead of creating a PowerPoint slide deck — a static artifact that quickly becomes outdated — you access the data warehouse to collect specific insights when you need them. This interaction with a knowledge base doesn’t require any particular skills since you ask questions using natural language. You collect this information in a dialogue format, just like you talk with your customers.

The Bright Future of User Research: People, Process, Tools was originally published in UX Planet on Medium, where people are continuing the conversation by highlighting and responding to this story.