Building a Growth Engineering Team That Tests Every Idea
Growth engineering sits where product value meets the work of getting users to that value. This guide draws on a conversation with Jason van der Merwe, Director of Growth Engineering at Strava, who was asked to build the company's growth team eight years before the interview; it has since grown to 70 people. The discussion covers how such a team is built, why testing is the engine, and how to keep the end user in view even when internal systems feel chaotic.
What makes growth engineering different
A growth engineer does not stop at delivering a spec. The role points toward the user and the business impact, which means a lot of learning happens along the way. Growth engineers ask how a flow can be improved, where friction can be removed, and whether a large change could shift the business. The test is whether the work actively improves something for the user rather than existing for its own sake.
The pace can feel uncomfortable to engineers who are used to longer, more stable delivery cycles. Growth work often involves assembling a short-term solution quickly to learn whether an idea deserves a bigger investment. Because the user stays at the center, this can require a willingness to bend the usual rules at times. That trait is useful in growth and less useful in infrastructure teams, where stability and predictability matter more.
The strongest growth engineers tend to be doggedly user-focused, and some of them could see themselves becoming product managers one day.
Managing chaos in an app business
An app business touches many channels at once: feature education, push notifications, email and more. The complexity can look like chaos, and even a successful app with strong organic growth is not immune. When Strava hired a Chief Design Officer, she asked for user mapping and the user journey. The team's honest answer was that they would like to see that too. Internally, it is sometimes hard to know exactly what every user is seeing at any moment.
Observability is a real challenge. The practical response is not to wait for perfect visibility before making decisions. Leaders who spend all their time keeping the team inside the lines may avoid downsides but miss upside opportunities. Jason points to Eric Schmidt's idea of innovation as managed chaos: everyone does their job while maintaining just enough order to make decisions.
Creating room to test every idea
Testing is one of the biggest tools for growth. New teams and individuals joining the growth organization are encouraged to spend their first few months running A/B tests. The point is to learn how to move from idea and hypothesis to testing and evaluation. A test does not have to knock the ball out of the park or push a metric that turns out to be important. What matters more is having a metric to aim for, thinking about it and brainstorming.
User research has been present since early on. Even as a small team, they ran user tests to help people move through onboarding. For Strava, a user's activation metric is uploading their first workout, along with following one person. The team noticed a drop-off on the recording screen because people did not know which button to tap. They labeled the buttons, then brainstormed and tested what language to use and where to put it. The word "go" turned out to be universal enough to work across even non-English speaking users. Jason describes this as one of the biggest wins in reducing friction to upload for new users, with a large increase in retention.
The 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, mostly leads spoke up, so the team moved to meetings by team and restored the productive testing discussions. Jason still reviews recent tests to question ideas and celebrate wins. His role, as he describes it, is to make it easy enough to test any and every idea. That approach is more inclusive and removes a natural bias toward defaulting to the loudest voice.
Feature education and user empathy
Many growth team members use the app every day. That is a gift in some ways, but it has a shadow side: you are not using the product from the perspective of a new user. Most Strava users do not run or cycle every day, so it can be hard for team members to put themselves in users' shoes. One way they do that is a toggle that switches a profile into "free mode." Much of the subscriptions team keeps that setting on all the time, which lets them see what those users see and identify questions newer users might have.
Feature education was not always a priority at Strava. In earlier phases, the team concentrated on adding capabilities for people who were already active in the app. Newcomers, and people returning after time away, were not the main audience for guidance about how the product worked. The turning point came when the growth team built an internal service for delivering feature education, such as pop-ups and tool tips. That service started as a growth-team-only tool. In time, other teams and product managers recognized its value and began using it to add more in-product guidance. One team is now testing the removal of feature education to see the overall impact on users.
Why it always comes back to the end user
If a growth engineer thinks like a product manager, keeping users front and center comes naturally. You become eager to test new ideas to remove friction. Feature education becomes essential because the aim is to be relentlessly helpful to users. Even managing chaos becomes less of a slog because a better product and more innovative ideas are on the other side.
A user-first habit shows up in each interaction a person has with the product, not only in the big launches. When a team keeps the user at the center of every decision, that orientation carries through to each touchpoint. In the effort to turn new installs into lasting, loyal users, the app that consistently puts the user first comes out ahead.
Frequently asked questions
What is the difference between a growth engineer and a product engineer?
A growth engineer moves toward the user and the business impact rather than only delivering a specified feature. The role involves rapid short-term solutions to test whether an idea deserves a larger investment, and it suits engineers who are strongly user-focused and comfortable with a faster pace.
What should a new growth team member do first?
Jason encourages new teams and individuals to spend their first few months running A/B tests. The goal is to learn how to go from idea and hypothesis to testing and evaluation, not to produce a perfect result on every test.
What does the Strava activation example show?
Strava defines activation as uploading a first workout and following one person. The team found a drop-off on the recording screen because users did not know which button to tap. Labeling the buttons and testing the wording and placement reduced friction, and the word "go" worked across even non-English speaking users.
What does managed chaos mean for an app team?
Managed chaos means accepting that an app business touches many channels and that internal visibility is rarely perfect. The team keeps just enough order to make decisions and move forward rather than waiting for complete observability.