A lot of product teams claim to be focused on their users. They might even have regular steps in their processes that remind them to put their users’ needs first.
But even if this is the case, there’s always more they can do to improve. This is why Teresa talks about continuous discovery in terms of forming new habits. It’s not so much about becoming perfect at continuous discovery (because there really isn’t such a thing). It’s more about looking for new ways to collect insights from users, uncover underlying assumptions, and explore the opportunity space.
This was the situation Sören Weber and the Search & Flow team at trivago found themselves in. They already had an established product and regular rituals that helped them consider their users. But there was still more they could be doing. Through a combination of leadership buy-in, reading, and coaching, they committed to mapping opportunities and testing assumptions before jumping to solutions and found new ways to truly put their users first.
Read on to learn the steps Sören and the Search & Flow team took to adopt continuous discovery habits at trivago.
Have a Product in Practice story you’d like to share with Product Talk readers? You can submit your story here.
Meet the Continuous Discovery Champions, Sören Weber and the Search & Flow Team
Our continuous discovery champion is Sören Weber, a Senior Product Manager on the Search & Flow team at trivago. Sören is quick to point out that this wasn’t a solo effort, though. He applied the continuous discovery habits with a cross-functional product team comprising a product manager (Sören), product designers Emilio Martins and Gabriel Hourigan, user researcher Mara Zocco, and the engineers who were all involved in the discovery activities as well.
trivago’s vision is to create the best hotel comparison website in the world and the Search & Flow team focuses specifically on the upper part of the funnel when users start their search. Their goals are to learn about user intent through user input and implicit signals and use the knowledge they gain to tailor the search experience and help the user move down the funnel.
Sören describes trivago as a mature product that found product/market fit 15 years ago with its hotel metasearch. Since then, they’ve deliberately focused on optimizing hotel metasearch to become the best comparison site for hotels. Sören says since this has been the focus, they’ve optimized the product a lot, iterated on all aspects of it, and run countless A/B tests. This means there are few obvious product improvements left and they need to dig deep to find improvement areas.
The inspiration to adopt continuous discovery came from different sources. Many members of the Search & Flow team were regular Product Talk readers. They had participated in the SVPG Inspired workshop, which helped build a foundation of continuous discovery. Then Marco Pretorius joined the company as Head of UX. He had worked with Product Talk during his time at Seera Group and got buy-in from the executive team to invest in continuous discovery and booked the Master Class for a large part of the product organization including product managers, designers, engineers, and user researchers.
The Challenge: Identifying and Prioritizing User Problems Consistently and Effectively
And while this was a good start, there was still room for improvement, says Sören, “We already had a focus on the user, but often didn’t go deep enough to properly understand the problem space and sometimes jumped to solutions too quickly.” Solution ideas came from different sources like stakeholders, foundational user research, data analysis, and inspiration from competitors.
The bias toward solutions was also due to a data-driven product approach and a strong A/B testing capability. This provided an “easy path” for A/B testing any time they weren’t sure whether something was a good idea. “This is obviously a costly approach and took a lot of time and effort for each iteration,” says Sören. Then if an idea didn’t work, it was hard to understand why and make a decision about what to do next. And since they didn’t linger in the problem space for too long, it was more common to come up with obvious ideas and much more difficult to come up with truly innovative ones.
Describing their old approach, Sören says, “We used to do a lot of solution testing, which was in our comfort zone, but it meant we had a limited understanding of the problem. This led to the fact that we worked on one solution idea for a long time without reaching a point to qualify it. We also relied a lot on usability testing, which evaluates a specific solution, but we didn’t know that much about how this solves a particular problem.”
Sören says feedback from outside the Search & Flow team also confirmed some of their areas for improvement. “We recognized that we should get a deeper understanding of the problem space, and started leveraging JTBD and combining foundational research insights with in-product UX measurement.”
First Steps in Mapping out Opportunities
As they developed more understanding of continuous discovery, the Search & Flow team began discussing how they could focus more on the problem space. Initially, they still believed that understanding the problem was a lengthy process that required extensive user research, which would take weeks to complete. “The continuous discovery habits mindset and the simple structure of the opportunity solution tree felt like a good middle ground,” says Sören. “We can focus on the problem space, but it still feels quick and pragmatic.”
The continuous discovery habits mindset and the simple structure of the opportunity solution tree felt like a good middle ground. We can focus on the problem space, but it still feels quick and pragmatic. – Tweet This
The Master Class provided the kick-start they needed to begin mapping out their opportunities. And this step proved to be a key point in trivago’s continuous discovery journey. When the product team began to focus more on opportunities, they changed their mindset to really understand the problem first.
Now that they’ve made the deliberate decision to try continuous discovery and become more outcome oriented, they do a quarterly product outcome exercise, conduct one to three user interviews a week, hold a product trio meeting each week to extract opportunities from interviews and update the tree, and they use the opportunity solution tree as a central map of what they’re doing and why. “We’re now thinking about the user’s needs first,” says Sören. And while this initially felt like adding more work, that’s no longer the case. “We made this our normal way of working and it’s helped us better understand the problem space.”
With continuous discovery, we’re now thinking about the user’s needs first. We made this our normal way of working and it’s helped us better understand the problem space. – Tweet This
Some Early Challenges and Obstacles
Learning how to map opportunities was a gradual process, not something that happened instantly. Sören says finding a good structure for their opportunity solution tree was a challenge. “We had countless iterations.” Discussing the tree both in the trio and with additional stakeholders helped them refine it. They also abandoned the idea of perfection: “We realized that there can’t be a perfect opportunity solution tree, but as long as it’s helpful, it’s fine.”
There’s no such thing as a perfect opportunity solution tree. But as long as it’s helpful, it’s fine. – Tweet This
In the beginning, it was difficult to see how to integrate continuous discovery habits into the existing landscape of product practices, especially with the Jobs To Be Done (JTBD) framework. “It is still an ongoing process, but we’ve realized that the JTBD map is a very good starting point for opportunity mapping and that both frameworks reinforce each other. Insights about user jobs and needs inform our opportunity solution tree and insights from the continuous interviews can help refine the understanding of the jobs to be done.”
Uncovering the opportunities that came out of customer interviews was another challenge initially. Interviewing customers to identify opportunities is a skill that needs to be developed. “Users don’t usually express their needs directly. You need to extract them from the story, which requires practice,” says Sören. “We needed to level-up our interview skills, which we did with the help of experienced interviewers in the user research team.” The interview guidelines by Glauco Cavalheiro below offer an overview of how they now run interviews.
Users don’t usually express their needs directly. You need to extract them from the story, which requires practice. – Tweet This
They also discovered that after a couple of interviews they weren’t identifying new opportunities anymore. Asking users about a past experience seemed to reach a saturation point where the same topics kept coming up over and over again. At this stage, they made two changes: They began asking more follow-up questions to really drill down into what a user said and they started giving the user tasks to demonstrate, like showing how they found out if an accommodation was family friendly on trivago or a competitor’s website.
How Opportunity Mapping Transformed Discovery at trivago
They began to identify user needs they didn’t know about before. For example, one product outcome was to increase the number of meaningful interactions (e.g. applying a filter or opening details about a hotel). One branch of opportunities and sub-opportunities they’d identified included: “I need to decide where to stay,” “I wish it would take less effort to find hotels that meet my preferences,” “I only want to see hotels that meet my preferences, not all the other irrelevant ones,” and “I need to trust that my preferences are reflected correctly.”
From user interviews, they learned that users didn’t trust the filters to narrow down the list of hotels, so they would spend a lot of time and effort going through dozens of accommodation options themselves to verify whether they met their criteria (such as having a pool or being pet-friendly)—even though trivago could filter those criteria for them.
Identifying this opportunity has led the team to come up with new solutions they hadn’t considered previously, such as highlighting what they filtered for in hotel details, showing reviews that include something related to the filter, or suggesting filters based on the search context. They’re still testing solution ideas, but uncovering new user needs like this has been an exciting change.
Prioritizing and handling external requests became much easier. “We still had a lot of requests directly from different stakeholders, but with the opportunity solution tree, we could have a much better discussion on whether their proposal actually helped to reach our desired outcome faster,” says Sören. The structure of the tree meant they could make compare and contrast decisions and stakeholders welcomed the opportunity solution tree because it provided a structure for product decision-making. Plus, connecting the product outcome to a business outcome was very helpful for getting buy-in and facilitating more strategic discussions.
The consistency of the opportunity solution tree didn’t just improve communication with stakeholders and the leadership team—it also had an impact within the product organization. Sören explains, “We currently have five product squads and once we started to use the same language and format for our opportunity space, it became easier to align the product teams.” They now do a regular opportunity solution tree run-through for every team where each product manager presents their tree and what’s changed since the last meeting based on key learnings from user interviews and assumption testing.
Quarterly planning is now more efficient. “It used to be a bit of a messy process to decide on OKRs and priorities, and now that we base the quarterly planning on the opportunity solution tree, it is much easier to communicate why a certain objective has been chosen,” says Sören.
We used to have a messy process to decide on OKRs and priorities. Now that we base quarterly planning on the opportunity solution tree, it’s much easier to communicate why a certain objective was chosen. – Tweet This
And finally, there’s less waste. “We had solution ideas we were planning to implement, but after mapping them in the opportunity solution tree, we decided not to pursue them,” says Sören. Now they always make sure solutions map to clear opportunities before moving forward.
In addition to mapping solutions to opportunities, Sören says assumption testing is another important part of the process that helps reduce waste and de-risk opportunities coming out of interviews. They need to gather at least a small amount of evidence to support their assumptions before moving forward with them.
What They’ve Learned and Advice for Other Teams
Reflecting on this journey, Sören offers a few key lessons and takeaways for others.
- Commit and stick to it long enough
Don’t expect the results to be immediate. It takes time to build the habit and get everyone bought into this way of working. “In the beginning, it felt like an additional burden before we saw the benefits,” says Sören.
- Get buy-in from your product organization
The adoption of the continuous discovery habits takes time and effort and the benefits might not be obvious in the beginning. This means it’s important to have the support from the rest of the organization. “In our case, it helped us a lot to have buy-in from the UX Research team run by Ruben Stegbauer and the product leadership to embark on the continuous discovery journey,” says Sören. He mentions specifically the support they received from product leads Raquel Santos and Nico Henrich, design lead Daniel Roux, Head of UX Marco Pretorius, and Head of Product James Neaves.
- Embrace the messiness and the fact that it’s hard
This is where the value lies, says Sören. “Product management is no perfect science. With the opportunity solution tree you can make sense of it and give it structure only to a point.” There will still be uncertainty and the messy “art” part, but your work will be much easier with the opportunity solution tree as a basis than without it.
Are you trying to implement continuous discovery in your organization? We regularly share success stories, challenges, tips, and tricks in the Continuous Discovery Habits community. Come join us!