The Growth Process

We have codified this internally to help other teams at Atlassian understand how to approach building product using a growth mindset. Hopefully, you find this helpful in building a growth competency within your team.

What is the Growth Process?

The Growth Process is a framework for growth teams that provides a set of guidelines for:

• Understanding which opportunities will have the greatest impact
• How to generate experiment ideas that have a high success rate
• How to validate that your ideas really do have a positive impact
• How to interpret and iterate on experiment results

Ultimately this process is successful when you start to see your metrics move in a positive direction. Teams should be generating more successful experiments (and corresponding product changes) because these guidelines maximize their likelihood of success.

Who is this process for?

Growth typically refers to a focus on increasing a business metric using experimentation. When a product has a well-defined goal, the Growth process should be applied because:

• The impact of changes are directly measured and provable. Because experimentation is used, you will have know how your changes impacted the customer. Intuition is proven or disproved.
• It’s the best way for understanding how to optimize a journey. With funnels and other data you know exactly where your effort will have the greatest impact.
• Data tells us where the biggest opportunities are, even if that work is outside our areas of expertise or ownership. Moreover, once these problems are apparent, they are often times solved with relatively simple solutions.

How is the Growth Process built?

The Growth Process consists of several broad stages:

• Define initiative – Defining your goals and researching opportunities as a team are completed prior to writing any code. It typically takes 1 to 4 weeks to complete this work, depending on the complexity and size of the problem you’re tackling.
• Envision solutions – During the ideation phase you will come up with experiment ideas that you believe will move your metric. This stage is brief; less than a week, and generally the entire team is involved.
• Run experiments – Preparing and developing a running experiment can vary by initiative, ideally you want to invest the least amount of time without watering down the hypothesis. Developers do and drive this work.
• Results – Analyzing results and deciding next steps usually takes about 1 week, and generally the entire team is involved. The most important step is to communicate your learnings to the broader team, regardless if the experiment was a success or not.