What is product discovery?
Product discovery is the process of understanding customer problems and testing possible solutions so a team knows what is worth building before it builds it.
Three questions sit underneath it: is the problem real, do people want this particular solution, and can you actually deliver it. Teams that treat discovery as a phase that ends when development starts tend to rediscover the same things later, at a worse moment.
Marty Cagan of the Silicon Valley Product Group, author of Inspired, describes discovery as reducing four risks before delivery starts:
- Value: will customers buy it or choose to use it?
- Usability: can they figure out how to use it?
- Feasibility: can engineers build it with the time, skills and technology they have?
- Business viability: does it work for the rest of the business, including sales, legal and finance?
What is the difference between product discovery and product delivery?
Discovery decides what to build, and delivery builds and ships it. Discovery produces evidence and tested ideas. Delivery produces working software. Most product teams run both at the same time, a practice called dual-track agile: while engineers deliver this sprint's work, the product manager, a designer and a lead engineer test the ideas for the sprints after it.
What are the stages of product discovery?
Product discovery moves through a problem space and then a solution space, and most teams break that into six stages.
- Set the outcome. Agree which customer or business result the team is trying to move, for example the share of new accounts that are active after one week.
- Explore the problem. Interview customers, read feedback and support conversations, and look at usage data to find where people get stuck.
- Map the opportunities. Write each unmet need or pain point down as an opportunity, then pick the one that matters most for the outcome.
- Generate solutions. Come up with several ways to address that opportunity before committing to one.
- Test the assumptions. Use prototypes, short experiments and usability tests to check the riskiest assumption behind each solution.
- Decide and hand over. Pick a solution, write down the evidence behind it, and move it into delivery.
The first three stages sit in the problem space and the last three in the solution space. Teams that skip the problem space usually end up testing a solution to a problem nobody has.
What is continuous discovery?
Continuous discovery is product discovery done every week instead of as a phase before a project. Teresa Torres, who popularised the approach in her 2021 book Continuous Discovery Habits, describes it as weekly touchpoints with customers, by the team building the product, where they run small research activities in pursuit of a desired outcome.
Torres calls that team the product trio: a product manager, a designer and a software engineer. The weekly rhythm matters because a team that talks to customers once a quarter makes most of its decisions in between, without them.
What is an opportunity solution tree?
An opportunity solution tree is a diagram that connects a desired outcome to the customer opportunities that could move it, the solutions for each opportunity, and the tests for each solution. Teresa Torres introduced it on her blog, Product Talk.
- Outcome at the top, for example "more trial teams invite a second colleague".
- Opportunities below it: customer needs, pain points and wishes, taken from interviews and feedback.
- Solutions under each opportunity, usually several per opportunity.
- Assumption tests under each solution, the cheapest way to find out whether it works.
Anyone can follow a solution up the tree to its opportunity and outcome, which makes the reasoning behind a roadmap item visible.
What techniques are used in product discovery?
The most common product discovery techniques are customer interviews, feedback analysis, prototype tests and small experiments, and each one answers a different question.
- Customer interviews about recent behaviour: what the person did last time, which tells you more than what they say they would like.
- Feedback and support analysis counts how often a problem comes up and in which segment.
- Surveys size a problem across many users once interviews have found it.
- Customer journey mapping shows the full experience and the steps where people drop out.
- Prototype and usability tests check whether people understand a solution and can use it.
- Fake-door and A/B tests measure real demand before the full feature exists.
- Competitive analysis gives context on what customers already compare you with.
Which frameworks are used for product discovery?
Teams usually combine one framework that structures the work with one that helps them choose between options.
- Double Diamond (UK Design Council): diverge and converge twice, first on the problem, then on the solution.
- Opportunity solution tree (Teresa Torres): outcome, opportunities, solutions and tests in one diagram.
- Lean Startup (Eric Ries): build, measure, learn, using the smallest experiment that answers the question.
- Jobs to be Done: describe the progress a customer is trying to make instead of the feature they ask for.
- RICE and Kano: RICE, from Intercom, scores reach, impact, confidence and effort. The Kano model, from Noriaki Kano, sorts features by how they affect satisfaction.
What is the difference between product discovery and customer discovery?
Product discovery and customer discovery get used interchangeably and they are not the same. Customer discovery asks whether the problem is worth solving at all. Product discovery starts once you accept that it is, and tests whether your answer to it holds up.
Customer discovery comes from Steve Blank's customer development method, written for startups that are still searching for a market. Product discovery is the ongoing version inside a company that already has customers.
What is an example of product discovery?
A typical example starts with a pattern in feedback and ends with a tested change. Say feedback about your product keeps mentioning "sharing reports". Interviews show that people share reports with managers who have no account, and they paste screenshots into emails to do it. The team writes the opportunity "managers without an account can't follow progress", sketches three solutions, and tests a clickable prototype of view-only links with eight customers. Six of them use it without help, so view-only links move into delivery with the interview notes attached.
Product discovery for enterprise product teams
At enterprise scale, discovery evidence is spread across many teams. Customer success hears one thing, sales calls surface another, and each product manager runs their own interviews. Without one shared place for that evidence, two teams can research the same problem in the same quarter and reach different conclusions.
A shared record fixes that. Every opportunity links to the interviews, tickets and feedback behind it, with a count and a source for each, so a roadmap review can question the evidence directly. A year later, a new product manager can still see why the decision was made.
How Usersnap supports product discovery
Usersnap brings discovery evidence into one workspace and keeps every item tied to its source, from the first comment to the scored opportunity.
- Calls and conversations: Channels imports Zoom Cloud call transcripts and closed Intercom conversations next to in-app feedback and micro-survey answers.
- Themes: Sentiment Sensor tags each new item with a topic and a sentiment, and shows the top positive and negative topics with the items behind each one.
- Hypotheses: the Hypothesis Generator suggests a likely root cause for a negative trend, with the supporting feedback, and the Solution Generator proposes a fix you can review.
- Prototypes: Usersnap turns the evidence into a prototype prompt and hands it to Alloy, which generates the visual concepts.
- Opportunities: the Opportunities Board scores each opportunity on value and effort, with scoring criteria you can customise, and syncs both ways with Jira and Asana.
- Your AI assistant: with the MCP connector, ChatGPT, Claude or Cursor can query your live workspace with your own permissions and search, create or update opportunities. The connector sees new data, and Customer Success can index older items.
Sentiment Sensor and the Hypothesis and Solution Generators are included on Professional plans and above, and available as an add-on on Starter and Growth. Channels is included from the Growth plan. To set this up from a template, see AI Ingestion Product Discovery.
Product discovery FAQ
Who is responsible for product discovery?
The product manager owns it, but discovery works best as shared work between the product manager, a designer and an engineer, the group Teresa Torres calls the product trio. With an engineer involved early, feasibility problems show up before an idea reaches the roadmap.
How long does product discovery take?
For a single feature, anywhere from a few days to a few weeks. Continuous discovery has no end date, because the team keeps a weekly rhythm of customer conversations and small tests alongside delivery.
What is the difference between Jira and Jira Product Discovery?
Jira tracks delivery work such as epics, stories and bugs. Jira Product Discovery is a separate Atlassian product for collecting ideas and evidence and prioritising them before they turn into Jira work items.
Can Usersnap generate hypotheses from customer feedback?
Yes. On Professional plans and above, the Hypothesis Generator suggests a likely root cause for a negative sentiment trend and shows the feedback items it is based on. You can then turn the hypothesis into an opportunity.
Related terms
See how product teams keep discovery evidence tied to every roadmap decision.
.svg)
