Building an engineering growth team and encouraging them to test

Strava’s Jason van der Merwe shares how testing is key to growth and why creating space for every idea drives better decisions.

Peter Meinertzhagen

Peter Meinertzhagen

Published

What do you get when you join expanding product value for your users with connecting users to that value? Growth engineering. Many larger companies are no strangers to this concept, but growth engineering is also gaining traction with app companies as well.

Jason van der Merwe, Director of Growth Engineering at fitness tracking app Strava, was tasked with building a growth team eight years ago. In that time, the team has grown to 70 people. Jason’s expansive team embraces two primary ideologies: constantly test ideas and focus on the user.

“I want our engineers to think like product managers — thinking about that user problem they’re trying to solve,” says Jason.

In a conversation on the Sub Club podcast, Jason discusses the mindset shift of growth engineering, managing chaos, and encouraging experiments that strengthen Strava day in and day out.

Growth engineering: beyond delivering as we’re told

According to Jason, growth engineering requires a shift in mindset away from solely delivering something you’ve been told to deliver. Instead, growth engineers move directly toward the user and the business impact.

“I think that orientation is really important because in order to understand and get to that [business] impact, you have to do a lot of learning,” Jason explains. “That learning comes in all different shapes and sizes.”

Growth engineers focus on optimization in many forms. How can we improve this flow? How can we reduce friction to this flow? How do we build something massive to really change the game for our business? All of these questions seek answers that will actively improve things for the user, rather than building something just for the sake of it.

Non-growth engineers might not be comfortable with the pace of play that comes with growth engineering at times. Growth engineers often need to “hack” together short-term solutions quickly to understand whether an idea is worth a bigger, long-term investment. 

Because the commitment to the user stays at the forefront, growth engineering requires a change in thinking that Jason calls “a willingness to almost break the rules at times.” (This is also the trait that likely makes growth engineers the wrong people to run infrastructure teams, Jason points out.)

Who makes the best growth engineers? Engineers who are doggedly user-focused, particularly those who could see themselves becoming product managers one day. 

Chaos in an app business is unavoidable but manageable

Building and managing an app business means keeping track of countless channels, from feature education to push notifications to email. And that complexity, more often than not, can look like chaos. 

Case in point: When Strava hired its Chief Design Officer, Anita Patwardhan Butler, earlier this year, she asked Jason for user mapping and wanted to see the user journey. The team’s response? “We’d love to see that, too,” Jason says.

The many layers of the app world mean that internal chaos simply comes with the territory, even for an app as successful — and with as much organic growth — as Strava. “It’s easy for people to look at Strava and [think], Oh, they have their stuff together, right? Clearly they know what they’re doing,” Jason says. “And internally, sometimes we don’t even know exactly what every user is seeing at any time.”

Observability is a major challenge. But the key to rising to the challenge is the willingness to move forward and make decisions even without perfect observability.

Leaders who spend all their time ensuring their team colors inside the lines — hoping to cut off downsides — are likely to miss a lot of upside opportunities. Holding back from innovation to avoid upsetting anyone might work for major, longstanding organizations, but it won’t work for those in the app space.

As a leader, Jason subscribes to a management theory of managed chaos coined by former Google CEO Eric Schmidt, who says that innovation is managed chaos. Within this framework, everyone does their job with the goal of maintaining just enough order within the chaos to make decisions. 

Creating the space to test every idea: the key to growth

Jason’s experience of starting Strava’s growth team has given him on-the-ground insights into what works, and one of the biggest tools for success is testing.

When new teams or individuals join the growth organization at Strava, Jason encourages them to spend their first few months running A/B tests. 

He wants them to figure out how to go from idea and hypothesis to testing and evaluation. “It’s okay if you didn’t knock the ball out of the park or the metric you were pushing for wasn’t that important,” he tells them. “What matters more is you have a metric you’re going for, you’re thinking about it, [and] you’re brainstorming.”

Another key to the success of the growth team has been the presence of user research since early on. Even as a small team, they were running user tests to help users move through the onboarding flow. 

For Strava, a user’s activation metric is uploading their first workout (along with following one person). But they noticed a dropoff on the recording screen because people didn’t know which button to tap. The team decided to label the buttons, then brainstormed and tested what language to use and where to put it. (It turned out that “go” was a universal enough word to meet the need across even non-English speaking users.)

“That was one of the biggest wins we’ve ever had in reducing the friction to upload… for new users, and that results in a huge increase in retention,” says Jason. 

Strava’s growth team also created a weekly meeting called “Experiments Weekly” to discuss what everyone was testing and exchange hypotheses and ideas. As the growth organization grew to 70 people, though, primarily leads spoke up in the meeting. In response, the team moved to meeting by team, reinstating the productive testing discussions that fuel Strava’s growth.

Even though Experiments Weekly is no longer a single collective meeting, Jason still takes time to review recent tests so he can question ideas and celebrate wins with his growth team.

“I think my role at Strava is to make it easy enough to test any and every idea,” Jason explains. He says this approach is more inclusive — moving fast and testing often allows the growth team to remove a natural bias toward defaulting to the loudest voice. Instead, welcoming everyone’s ideas allows you to truly find the best ones.

“We want to test things, and we want to create safe spaces to do that. Some amazing stuff comes out of that,” Jason says.

Tapping into user empathy through feature education

Many members of Strava’s growth team use the app day in and day out. 

In some ways, that’s a gift. However, there’s a shadow side to that reality as well. “You’re not using the product from the perspective of a new user,” Jason explains. “That kind of empathy is tough.” 

Most Strava users don’t run or cycle every day, so it can be difficult for Strava team members to put themselves in users’ shoes. One way they do that, though, is with a toggle that allows them to switch their profile into “free mode.” Jason says that much of the subscriptions team keeps that setting on all the time, which allows them to see what those users see and identify questions newer users might have. 

Strava didn’t always prioritize feature education. Instead, teams focused on building features for engaged users, rather than new or lapsed users who needed to be taught how to use the product.

But that changed when the growth team created an internal service allowing them to implement feature education, including pop-ups and tool tips. At first, the service was used exclusively by the growth team. Over time, though, multiple teams and product managers saw the importance of using the service to offer more feature education for users.

Funny enough, one team is now doing testing where they remove Strava feature education to see the overall impact on users. “It was so wild to me because I tried for so long to get people to do feature education. Now we have too much feature education and we need to take a step back,” Jason says.

It always comes back to the end user

If you think like a product manager (as a growth engineer), keeping users front and center comes naturally — and the rest will follow. 

You’ll become eager to test new ideas to remove friction. Feature education becomes essential because your aim is to be relentlessly helpful to users. Even managing chaos turns into less of a slog because you know a better product and more innovative ideas are on the other side of it.

When you consistently prioritize users, that mindset comes through in every single touchpoint with your product. In the race to turn downloads into loyal fans, the user-focused app wins every time.


This article is based on an episode of our podcast, Sub Club, which explores best practices and insider secrets for scaling your app. Subscribe via Apple, Google, Spotify or wherever you get your podcasts.

In-App Subscriptions Made Easy

See why thousands of the world's tops apps use RevenueCat to power in-app purchases, analyze subscription data, and grow revenue on iOS, Android, and the web.

Related posts

TAM SAM SOM for apps: Everything subscription apps need to know
TAM SAM SOM for apps: Everything subscription apps need to know 
Growth

TAM SAM SOM for apps: Everything subscription apps need to know 

What are total addressable market, serviceable addressable market, and serviceable obtainable market and how should you use them?

Peter Meinertzhagen

Peter Meinertzhagen

April 11, 2024

How to make money on the visionOS App Store
How to make money on the visionOS App Store
Growth

How to make money on the visionOS App Store

It may not be the gold rush that the original App Store was, but now is the time for experimentation and innovation.

Charlie Chapman

Charlie Chapman

April 10, 2024

“Meta has a very strong algorithm… They’re good at finding your audience for you” — Marcus Burke, Independent Consultant
Growth

“Meta has a very strong algorithm… They’re good at finding your audience for you” — Marcus Burke, Independent Consultant

Marcus Burke on the past, present, and future of Meta ads.

David Barnard

David Barnard

April 3, 2024

Want to see how RevenueCat can help?

RevenueCat enables us to have one single source of truth for subscriptions and revenue data.

Olivier Lemarié, PhotoroomOlivier Lemarié, Photoroom
Read Case Study