Continuous Product Discovery: How to Make It a Weekly Habit

Great products are discovered, one customer conversation at a time.

Most teams have plenty of tools for shipping and almost none for deciding what to ship. Roadmaps fill up. Features go out. Then someone asks the quiet question in the retro: did any of this solve a real customer problem?

Continuous product discovery is the answer to that question, and it is calmer than it sounds. Product discovery coach Teresa Torres defines it as weekly touchpoints with customers, run by the team building the product, in pursuit of a desired outcome.

Torres has taught this approach to more than 7,000 product people through Product Talk Academy, and the through-line is simple: small and frequent beats big and rare. This guide covers what continuous discovery is, why it works, and how to set it up without blowing up your sprint.

TL;DR

• Continuous product discovery means weekly customer touchpoints run by the team building the product, not a one-time study at the start.
• It replaces project-based research, where teams learn once and then guess for the next six months.
• The core habit: a clear outcome, weekly interviews, and small assumption tests instead of big, rare studies.
• Structure the work with an opportunity solution tree, Torres's map of outcome, opportunities, solutions, and experiments (introduced in 2016).
• The whole product trio, product manager, designer, and engineer, shares the work, and every insight lands somewhere the team can search later.

Want the numbers behind the customer listening gap? HeyMarvin surveyed 309 research and product professionals for The State of Modern Research 2026: Closing the Listening Gap. The finding that stings: 94% of leaders say research should drive decisions, yet only 27% consistently use it. Continuous discovery is how you close that gap.

What continuous product discovery is

Continuous product discovery is a sustained practice of researching what to build, done in small pieces, all the time. Torres sets a clear bar for it. The minimum is weekly customer touchpoints by the team building the product, in pursuit of a desired outcome.

Break that definition into parts, and it gets practical:

• Weekly. A regular cadence that the team commits to.
• With customers. Real users who give you their actual feedback and reactions.
• By the team building the product. The people making decisions hear the evidence firsthand.
• A desired outcome. Discovery serves a goal, such as reducing churn, not a feature you already decided to ship.

The contrast is project-based discovery, where a team runs one study at the kickoff, then builds for months on assumptions that quietly go stale. The Interaction Design Foundation frames continuous discovery as research conducted as small, frequent activities across the whole product lifecycle, so customer feedback informs every decision rather than one meeting in week one. If you want a wider view of the sibling methods, continuous discovery research and the full set of product discovery techniques sit alongside this habit.

Why continuous discovery beats one-and-done research

Continuous discovery keeps the roadmap current. The team talks to customers every week. Assumptions get tested while they're still cheap to change, before they harden into a shipped feature. That steady input does three things:

1. It kills bad ideas early. You learn a solution misses the mark in a 30-minute interview, not after a quarter of engineering.
2. It builds shared conviction. When the whole team hears the same customer story, roadmap debates turn into evidence. That shift is a big reason user research for product managers has moved from nice-to-have to core skill.
3. It compounds. Each week of learning adds to the last, so the team's model of the customer gets sharper over time instead of resetting at every project.

There is a catch, and it’s the reason many teams stall. Weekly interviews only pay off if the team can find and reuse what it learned. Insights scattered across recordings, docs, and Slack threads are insights lost.

Weekly interviews only pay off if the team can find what they learned. When notes live in five tools and three people's heads, discovery stalls by Thursday. See how Marvin keeps every interview, tag, and theme searchable in one place, and book a demo.

How to set up continuous product discovery

You don’t need a research budget or new headcount to start. You need a cadence and structure. Here’s a step-by-step way to build the habit.

Step 1 - Start with a clear product outcome

Discovery without a target becomes trivia. Pick one outcome the team is trying to move, such as "increase weekly active teams" or "reduce first-week churn." Frame it as a customer behavior or a business result, not a feature. Every interview and test then serves that goal, which keeps the work focused and easy to defend.

Step 2 - Book weekly customer touchpoints

Put recurring interviews on the calendar and automate recruiting so you never start from zero each week. One or two conversations a week is enough to begin. Keep them short and behavioral: ask about specific past events ("walk me through the last time you..."), rather than hypotheticals. If you’re new to the format, these tips for running user interviews will keep the sessions honest and useful. The point is rhythm. A standing slot beats a heroic sprint of 10 interviews once a year.

Step 3 - Map the problem space with an opportunity solution tree

Torres introduced the opportunity solution tree as a visual map for discovery. It has four layers:

• Desired outcome at the top, the goal you are trying to move
• Opportunities, the customer needs, pains, and desires that drive it
• Solutions, the candidate ideas for each opportunity
• Experiments, the assumption tests that check whether a solution works

Torres described it as "a simple way of visually representing how you plan to reach the desired outcome." The tree forces the team to explore the problem space before jumping to features. You compare and contrast options instead of arguing whether to build one pet idea.

Step 4 - Run small assumption tests, not big studies

Under each solution sit assumptions that must be true for it to work. Test the riskiest ones with the smallest experiment that gives a signal: a fake-door test, a prototype walkthrough, a quick survey, a landing page. Small tests are fast and cheap, which is the whole point.

Step 5 - Make the whole product trio share the work

Discovery is not a solo research job. Torres and the IxDF both stress the "product trio," the product manager, designer, and engineer, participating together. When all three hear customers directly, decisions stop needing a game of telephone. Insights don't get watered down as they pass from one person to the next. Rotate who runs sessions so the habit does not rest on one person.

Step 6 - Store what you learn where the team can find it

This is the step teams skip, and the one that makes the rest stick. Every interview, tag, and theme needs a home the whole org can search, or the same study gets run twice and hard-won findings evaporate. A shared research repository with a consistent tagging system turns a pile of weekly notes into a growing, reusable body of knowledge. Insights compound only when you can find them.

Common mistakes teams make

A few failure modes show up again and again:

• Treating it as a phase. Continuous discovery is a habit, not a stage that ends when delivery starts. If it stops the moment a sprint gets busy, it was never continuous.
• Interviewing without an outcome. Conversations with no target produce interesting quotes and no decisions.
• Jumping straight to solutions. Skipping the opportunity layer means the team optimizes features nobody asked for.
• Hoarding insights. Findings locked in one person's notebook help no one. Shared and searchable is the standard.
• Going big and rare. A single massive study each quarter is the opposite of the practice. Small and weekly wins.

Frequently asked questions (FAQs)

Here are the questions product and research teams ask most often when they start with continuous discovery.

How is continuous discovery different from regular product discovery?

Traditional product discovery often happens once, at the start of a project, and then the team builds on those findings for months. Continuous discovery makes research an ongoing habit, with small, frequent customer touchpoints across the whole product lifecycle, so decisions stay grounded in current evidence.

How often should a team talk to customers?

Teresa Torres sets the floor at weekly touchpoints by the team building the product. One or two short interviews a week is a realistic start. Consistency matters more than volume, so a steady cadence beats an occasional burst of research.

Who should run continuous discovery?

The product trio, the product manager, designer, and engineer, shares the work rather than handing it to a lone researcher. When the people making decisions hear customers firsthand, findings lose less in translation. A dedicated researcher, where one exists, can raise the quality and set standards.

What is an opportunity solution tree?

An opportunity solution tree is a visual map Teresa Torres introduced in 2016 with four layers: a desired outcome at the root, the customer opportunities that drive it, candidate solutions, and the experiments that test those solutions. It helps a team explore the problem space and make decisions as compare-and-contrast.

Can small teams do continuous discovery?

Yes. The practice scales down well because it favors small, frequent activities over large studies. A team of one PM and one designer can book a couple of interviews a week, keep an opportunity solution tree, and run quick assumption tests. The habit matters more than the headcount.

Turn discovery into a habit that sticks

Continuous product discovery isn’t a heavier version of research. It’s a lighter, steadier solution with a clear outcome, weekly conversations, a map of the problem space, and small tests that keep the team honest. Do it for a few weeks, and the roadmap debates change tone because answers come from customers instead of stakeholder opinions.

Ready to make discovery a habit, not a heroic effort? Marvin gives product trios one searchable home for interviews, themes, and cited answers, so weekly touchpoints compound instead of evaporate. Book a demo, and put customer input back at the center of every product decision.