01 The problem
Product discovery has excellent theory — the Mom Test, opportunity-solution trees, RICE prioritisation, assumption testing — scattered across books and blog posts. Knowing the frameworks and running them on a real problem are very different things.
When I was learning discovery myself, I kept hitting the same wall: I understood each framework in isolation but had no connective tissue between them. Where does a Mom-Test interview feed into? How does what I learn become a prioritised bet? The theory lives in fragments; the practice needs a flow.
Don't teach the frameworks again — sequence them. Discovery Funnel connects interview → insight → opportunity → prioritisation → experiment as one continuous workflow, so a learner (or a working PM) has a path to follow instead of a pile of concepts.
02 How it works
Mom-Test questions
Structured, bias-resistant interview questions that get at real behaviour, not flattery.
Opportunity tree
Insights organised visually into an opportunity-solution tree so structure emerges from the research.
RICE scoring
Opportunities scored on Reach, Impact, Confidence, Effort to force honest prioritisation.
Experiment tracker
Assumptions turned into trackable experiments, closing the loop from question to evidence.
03 The design decisions
- A funnel, not a toolbox. The value is the sequence. Plenty of tools give you a RICE calculator; the point here is that RICE comes after opportunity mapping, which comes after real interviews — the order is the teaching.
- Mom-Test questions built in. The most common discovery mistake is asking leading questions. Baking good questions into the tool removes that failure mode by default.
- Visual opportunity mapping. Trees are spatial; a list flattens the relationships that make an OST useful. The visual layout is doing real cognitive work.
04 What I'd watch next
Discovery Funnel is live but not yet validated with real users. Open questions:
- Do people actually follow the funnel end-to-end, or cherry-pick the one framework they already like? If they skip stages, the sequencing value isn't landing.
- Is it a learning tool or a working tool? It could serve PMs learning discovery, or PMs doing discovery. Those are different products; usage would tell me which to lean into.
- Where's the friction? A guided workflow can feel supportive or restrictive. Finding where users drop off would sharpen the flow.
05 What building it taught me
- Sequencing is a product. The frameworks aren't mine — but the decision to connect them into one flow is the actual design work, and it's what makes the tool useful.
- Building the tool taught me the craft. Turning discovery theory into a working funnel forced me to understand where each framework genuinely feeds the next — deeper than reading ever did.
- Guardrails beat instructions. Building the good questions in beats telling users "ask better questions" — the same principle as Tailor refusing to fabricate.