I was introduced to the term Continuous Product discovery in our monthly book club meeting at HungerStation where I am the Senior Product Manager for Grocery and Quick Commerce. Here we decide on a book to read and our choice of book was Continuous Product Discovery by Theresa Torres. We discuss our thoughts and learnings from the book on a on bi-weekly basis.

We decided to test out the above Discovery process with my Squad which was Grocery Q-commerce. The article below will help you to understand the discovery process step by step. It will also reflect on what went good, what went wrong and our key takeaways and learnings from the process.

Book club kicked-off

• We focus on defining the right KR and ambitious but achievable outcome.
• No output conversation for most of the OKR preparation.
• Higher flexibility of quarterly deliverables.
• Continuous Discovery Habits introduction.
• It looked like a framework that fits our needs.
• We started learning by doing

Our road to completing almost the whole tree –

The Opportunity Solution Tree is a visual representation of how you plan to achieve your desired outcome. The goal is to overcome many of the common pitfalls that occur during the product discovery phase by explicitly mapping out our opportunities and generating multiple ideas on how to achieve it.

Step 1: Identify the desired outcome: Narrow your goal to a single metric you want to improve (e.g., revenue, customer satisfaction, retention, etc.).

Step 2: Recognise opportunities that emerge from generative research: needs, pain points, desires.

Step 3: Brainstorm on solution: The caveat, however, is that a potential solution must directly link to an opportunity — otherwise, it’s just a distraction from the primary goal of your OST.

Step 4: Experiment to evaluate and evolve your solutions: Now, it’s time to test a single solution with sets of experiments.

1. Define the Outcome.

1. We planned our outcomes based on the our Quarterly OKRs
2. Align on which outcome we will focus on during the next couple of meetings.
3. Other outcomes should follow, but we do not want to lose focus right now.

2. Experience mapping.

1. When working on the outcome for the first time it might feel overwhelming where to start
2. It helps to map out what we already know.
3. Not to lose focus let’s set the scope of our experience map by creating an outcome based questions.

Example:

1. What’s preventing our customers from making the purchase?

2. What’s preventing us from fulfilling an order?

3. Mapping the opportunity space.

1. Single customer might elicit dozens of opportunities.
2. We need to structure the problem space before we can dive into solving it.
3. Our job is to address customer opportunities that drive the desired outcome.
4. We’ll use parents, child and siblings relationship in our tree
5. When we learn to think in the tree structure, it helps us deconstruct large problems into the series of smaller more solvable ones
6. We need every opportunity to be distinct from every other opportunity. Otherwise we’re going to work at all of them at once.
7. Use the steps from experience map as the “parent” opportunities. This will help you ensure there is no overlap between them.Define one opportunity for each moment in time in which that need, pain point or desire occurs

4. Take an inventory of the opportunity space.

1. Is this opportunity framed as a customer need, pain point or desire not a solution?
2. Is this opportunity unique to this customer, or have we seen it in more than one interview?
3. If we address this opportunity, will it drive our desired outcome?

If answer to all three questions is “Yes”, we want to add it to our OST

5. Prioritise opportunities.

1. Customers don’t care about outputs.
2. Customers care about solving their problem or a need.
3. We want to solve one opportunity at a time. It will allow us to explore multiple solutions.
4. We want to prioritise top-down
5. How to prioritise?

Opportunity sizing eg. How many customers are affected? How often?

Market factors eg. are there any trend on the market that might affect this opportunity?

Company factors eg. does this opportunity support our company vision and strategy?

Customer factors eg. how important it is to our customers?

6. Generate Ideas.

1. Let’s review our target opportunity to make sure everyone shares the common understanding.
2. We are going to generate ideas alone.
3. After that we’re going to move them into the sheet below
4. Let’s describe each one of them to ensure proper understanding.
5. For a short period let’s get back again to the individual work and generate few more.
6. After ideation happened let’s evaluate every ideas with simple question — “Does this idea solve the target opportunity?”
7. It’s okay if the solution is partially addressing. We want to weed out the ideas that don’t solve target opportunity at all.
8. While voting the only criteria should be how well the idea addresses the opportunity

6. Assumptions mapping.

1. We want to identify “Leap of faith” assumptions
2. These carry the most risk and thus need to be tested
3. We want to start with asking ourselves a question: How much do we know about this assumption?”
4. Do we already have some evidence?
5. Then we want to understand how important this assumption is to the success of an idea.
6. This exercise doesn’t need to be precise — we are mapping assumptions relative to each other.
7. Once we’ve completed the mapping we will start with testing the assumptions that are placed in the top right quadrant. Only 2–3 the most important.

Our reflection on the process –

What did we do Right?

1. Qualitative user research and identifying user pain points and mapping them to the user journey and translating them into the flow/funnel of the app.
2. The right thing to do after mapping on pain points is to pick a funnel in the app with maximum pain points and focusing on it to chalk out opportunities.
3. We mapped pain points for Discovery, Shopping Experience and Order fulfilment and further divided them into Journey and User objectives and decided to pick shopping experience as it fitted desired outcome the best
4. Defining parent opportunities mapping them together and prioritising them to generate ideas and assumptions

What went Wrong?

1. Continuous discovery process active is periodic user interactions/interviews which lead to a proper discovery of user pain points. We faltered here due various issues like — low turnout of users, language constraints.
2. We chalked out the tree with Product trio — PM, PD and EM and we often found ourselves during out brainstorming sessions focusing on solutions rather than opportunities.
3. We often found ourselves of killing ideas due to feasibility as a constraint leading to a limited set of ideas due to a constrained mindset.
4. Effort to set user interviews and redoing the OST cycle is a continuous journey and we often found ourselves stuck at not being able to move forward due to workload on other fronts.

Key takeaways

TIME & EFFORT

1. Building the first OST can take time. Set expectations before and be patient.
2. After initial tree, dedicate +/- 2 hrs a week to update it and plan further actions.
3. Create OST towards key metrics that you know you will follow throughout the year.

CUSTOMER INTERVIEWS

1. Without already mapped customer insights (qualitative & quantitative) it will be impossible to start.
2. Make sure you have a continuous flow of user insights to develop the tree.

FACILITATION

1. Establish Product-trio (PM <-> EM <-> Design) and share the responsibility of the tree. Having 1 leader might not be enough to drive the involvement. I had a great Combination here with Krzysztof Bondar, Magdalena Pitula-Szpakowska & Wesam AlYazjeen. The creation of the OST would not have been possible without this Awesome Team
2. If you have that option, include external facilitator eg. Agile Coach to introduce you to and support you with the technique.

MINDSET

1. Make sure you and your team get rid of output-driven mindset till the very late stages of OST
2. When you do the above, you start seeing needs, problems and desires of your customer base that can be addressed in multiple ways.