Continuous Discovery

Of all the external frameworks I have adopted, Teresa Torres' continuous discovery is the one that changed how I run a team day to day, not just how I think. Her definition, which I use almost verbatim: discovery is "at a minimum, weekly touchpoints with customers, by the team building the product, where they conduct small research activities in pursuit of a desired outcome." Every clause in that sentence is load-bearing, and most teams violate at least three of them.

#Unpacking Torres' definition

• "Weekly touchpoints" means not a quarterly research sprint. Cadence is the whole point. Insight from a once-a-quarter study is stale and half-forgotten by the time it matters, while weekly contact keeps the customer present in every decision.
• "By the team building the product" means not a research department that hands findings over a wall. When the people writing the code and the specs hear the customer directly, the insight survives translation. Outsourced research dies in a slide deck.
• "Small research activities" means not a giant study. A 20-minute interview, a quick prototype test. Small and frequent beats big and rare.
• "In pursuit of a desired outcome" means discovery is anchored to an outcome, not open-ended curiosity. This is what keeps it from becoming research theater, and it is why the Opportunity Solution Tree is its companion tool.

#The product trio

Torres argues discovery should be done by a product trio, the product manager, a designer, and an engineer, making decisions together. This is the part teams resist most and the part I defend hardest. The reasoning is that each role catches a different one of The Four Big Product Risks. The PM is attuned to value and viability, the designer to usability, the engineer to feasibility. Send only the PM to the interview and the team re-litigates everything later through a game of telephone. Send the trio and the people who will build the thing heard the customer's pain with their own ears.

In my experience the engineer in the room is the secret weapon. An engineer who has watched users struggle proposes better solutions and argues less about scope, because they are solving for the same problem you are instead of taking dictation.

#How this fits my model

The competency model names Voice of the Customer and Fluency with Data as competencies but is thin on the operating cadence that produces insight reliably. Continuous discovery is that cadence, the practice layer underneath the competencies. It is how a team stops being a feature factory reacting to a backlog and becomes one that is continuously learning its way to outcomes.

#Where I push back

I am a fan, not a zealot. Two honest caveats:

• The weekly bar is aspirational, not sacred. On platform teams whose "customer" is an internal team, or in regulated betting contexts where you cannot just DM users, strict weekly contact with end users is sometimes impossible. The spirit of it, regular and direct and small, matters more than literal Tuesday interviews.
• It can become its own kind of theater. A trio that interviews weekly but never ships a change has just moved the cargo cult from "we did research" to "we do continuous discovery". The test is whether the touchpoints are changing decisions. If the roadmap looks identical with or without them, you are performing discovery, not doing it.

#How AI is changing it

The biggest practical barrier to a weekly cadence was always the synthesis tax. Transcribing, coding, and writing up each touchpoint ate the time you would have spent on the next one. AI collapses that, transcribing and theming interviews in minutes, which makes a genuinely weekly habit feasible for a small trio for the first time. The risk is letting the model's tidy summary replace the team's own ears. The whole point of "by the team building the product" is the direct, messy exposure, and a clean AI digest can quietly undo it. See AI in Discovery and Research.

#Continue Reading

• The Opportunity Solution Tree for Torres' companion tool that keeps discovery anchored to an outcome.
• User Research and Discovery for the methods you run in each weekly touchpoint.
• The Four Big Product Risks for the four risks the product trio is built to catch.
• Voice of the Customer for the competency this cadence builds.
• Product Teams vs Feature Teams for the team model continuous discovery assumes.