Hi, I’m Tony Jones, and I’m a designer on Atlassian’s Growth team! In the first part of this series my teammate, Jet, talked about the 5 continuous discovery best practices my team implemented. In this article, I’ll dive into how we schedule, structure and report out on customer interviews to enable continuous discovery.
Glad you asked. Continuous discovery is about including customers throughout your decision-making process. The idea is that regularly hearing their stories will help you identify new opportunities, and close the gap between how you think your products are used and how your customers actually use them.
It’s nothing new. Marty Cagan at SVPG first started writing about continuous discovery more than 10 years ago, and we’ve been talking about it at Atlassian for a while too.
My team has found regularly chatting with customers to be one of the best parts of our week, and this article will walk you through how we set it up, because with a bit of a steer it can be easy and super valuable.
In the past, we aimed to do customer research as part of discovery, and again when measuring impact to understand why we saw the results we did. But in reality, this was hard to fit in, and it meant we rarely had input from customers while we were choosing a hypothesis and designing the solution.
It can be hard to imagine fitting in weekly customer conversations, I procrastinated on it for months, but a combination of three things made us try it:
✅ We could minimize the number of decisions we made without customer input
🔮 We could go beyond ideation sessions and identify new opportunities and pain points
🎛️ We could better understand how people actually use our products
That’s when I made the following customer call page, and we got started.
Here’s what a typical week looks like for us.
On Monday, we look at our product roadmap and chat about experiments we’re designing or kicking off so we can choose a theme to speak to customers about. For example, we had a couple of ideas coming up on helping customers discover how to make product roadmaps with Jira Product Discovery, so product roadmaps was our theme for the week.
After choosing a theme we create a guide broken into topics we want to discuss. Under each topic, we add 2-3 questions to prompt customers to share their experiences. We’ve found it helpful to have fewer questions, because it makes you feel less rushed and leaves room for follow-up questions.
With every discussion guide, we start with 3 background questions to understand what their company does, what their role is, and their satisfaction with the Atlassian products they use. This is a great invitation for customers to share top-of-mind pain points and opportunities.
We use User Interviews to find customers, and we recruit in 3-week blocks. This seems to work well as some customers are ready to speak that week, while others need more notice.
Every third week we duplicate our last User Interview project and add 4 sessions per week to cater to people who work in different timezones. We launch the project and set a maximum of 1 session per day to keep it manageable.
Come Tuesday morning we have a stack of candidates for interviews. When reviewing participants, we pay close attention to their responses to our screener questions about their role and day-to-day responsibilities - people with clear and detailed responses here are more likely to share clear and detailed responses in the interview.
We try to have 2 conversations per week. While this won’t give you a perfect saturation, it will help you make better decisions and identify new opportunities. We like to have chats on Wednesdays and Thursdays, because this gives us time to prepare on Tuesday if need be.
At the beginning of the customer conversation, we take 5 minutes to chat and connect. Building a small amount of rapport will help people relax and feel inclined to share more with you. Following the intro, you want the customer to be speaking 90% of the time. Be okay with silence; let people think – often the most useful responses come after someone has spent 10 seconds pulling faces and racking their brain!
We try to have 2 people in each session, one facilitating and focused on the conversation, and the other observing.
After the interview we quickly debrief, and import it to Dovetail using automation ⚡️ to get a summary. From the transcript, we grab quotes, tag interesting parts, and copy video highlights to add to our post-conversation write-up.
Easier said than done, but we aim to block out 60 minutes after the conversation to debrief and do the write-up and share out. Coming back to it the following day is tough, and you forget a lot of stuff. Better to do it while all the information is fresh.
We share our post-interview write-ups with our stakeholders in Slack. To make it something people can learn from at a glance, we include the summary and a short video highlight.
Lastly, we add insights to relevant ideas in our discovery project so we have qual and quant data together when we're ready to start this. We include a customer quote to bring it to life, a link to the summary in Confluence so we can find it, and a rating of its impact so we know how much it should inform our work.
Ready to go? Here are 4 things to keep in mind:
Hey @Stijn Veer_ van der I share every interview in related Slack channels with a handful of key insights and opportunities. I find most people glean these key points rather than read the full write-up, and that's cool because it's often something we reference again later.
Feedback has been positive - some people will @mention others if there is an insight relevant to them, or ask questions that I can then build into future customer conversations. I also see people referencing these write-ups and customer quotes in their project posters.
Most of the insights connect with work we have planned for later, I store the ones that don't in Dovetail and aim to review these when we do our quarterly plans.