Continuous discovery process for Kiwi.com

Process

Based on the insightful book Continuous Discovery by Teresa Torres, this approach to product discovery aims to create a structured and sustainable framework for deciding which features and strategies to develop. It leverages the vast amount of insights from various streams that constantly flow in. Rather than making independent bets based on assumptions, Torres' discovery approach continuously explores and develops opportunities to increase confidence in their impact before prioritizing them. This may involve additional research, expert reviews, or even testing.

However, not all companies are the same, as not all products are the same, so implementing a master blueprint and expect it to work just won't do the trick. If we combine that with the amount of stakeholders impacted by such a substantial change in phylosophy and ways of working, we find ourselves facing a significant strategic and communication challenge.

The preliminary work

Before defining my team's version of the framework, I conducted preliminary research, starting with a thorough reading and dissection of Teresa Torres' book. I also observed how other teams with similar approaches operated, both within and outside the company, and analyzed the documentation they produced.

Additionally, before any process design or action took place, I conducted a series of informal interviews with different team members. This helped me understand the specific challenges we were trying to solve and gather diverse perspectives to align on the definition of value. These exchanges were also aimed at establishing initial buy-in, especially from the rest of the leadership, as their support was paramount for the success of this effort.

Our continuous discovery process

The version of the continuous discovery process we piloted relied in 3 main pillars:

1. A core team, comprising representatives from product, design, and research, was responsible for several key tasks. They prepared insights for prioritization sessions, tracked the opportunities being discovered, initiated necessary discovery efforts, and facilitated and maintained both sessions and documentation.
2. We held weekly public discovery sessions where the entire team discussed and initially prioritized new potential opportunities. Additionally, these sessions involved ideating on already prioritized opportunities that were ready for solution development. These sessions were critical for generating value for both the team and the product, but required strong facilitation to ensure that the multidisciplinary team remained engaged and committed to the outcome of the process.
3. A detailed documentation was essential for capturing the state of each opportunity, providing a snapshot of our discovery efforts, and driving efficiency across the process. This documentation consisted of three main components: an inbox where initial insights arrived, a playground where prioritized opportunities were developed to build confidence, and an opportunity map that contained all relevant opportunities and linked them to the solutions designed to address them. These solutions were subsequently tracked in Jira, our primary tool for project management.  

The pilot strategy

As with any significant change, the introduction of this process required careful management.

Firstly, it needed thorough testing. More importantly, it impacted a diverse range of teams and individuals who already had established processes, expectations, and opinions. Each of us, from different angles, understood what needed improvement in our own way. Such situations, common in change management, create a breeding ground for pushback, lack of commitment, and potential failure of the new process.

For this reason, it was crucial to define a clear communication and introduction strategy to progressively demonstrate the value of the change to each team, beginning with minimal dedication requirements. In other words, the change's scope was gradually increased as its value became more apparent, with benefits communicated in ways that resonated with each team.

In our case, this meant utilizing a pilot quarter and, to keep the challenge manageable and maintain our agility, implementing it within the team where I had the most initial buy-in. During this period, the core team gradually expanded from a few trusted individuals to include more team members from different disciplines. The frequency of public sessions also increased, and their purpose evolved as attendees became more engaged. Additionally, the number of opportunities discussed grew. This progress was accompanied by full transparency through documentation, Jira, and reports during public sessions.

At the end of the quarter, once the roadmap for the following quarter was defined, I held a retrospective for the entire team to assess the change's value and define the next iterations, sharing metrics that summarized the results obtained. This session also aimed to bring more leadership members closer to the core team as I pushed to implement the new continuous discovery project across the other teams I supervised.

Results and follow-up

As cycles of the new continuous discovery process went on and stakeholder bought in, our opportunity map grew, and so did our Jira backlog. However, in contrast to what was happening before, our items, and the roadmaps they became part of, were concise, detailed, supported and easy to prioritize. This allowed for not only higher impact, but also for an easier and more relevant measurement of results.