From research to opportunities
After a few weeks of interviews and visits you have a pile of notes, photos and tickets. This lesson turns that pile into a short list of customer needs with evidence behind each one, an opportunity solution tree that links them to one outcome, and a ranked list of the assumptions to test first. It also shows where AI speeds this up and where it quietly misleads you.
By the end you can
- Turn snapshots, field notes and feedback into opportunities, counted by person and traced to sources
- Run a weekly always-on feedback loop from support tickets, app reviews and field chat groups
- Use AI for a first pass at tagging and clustering, and check it so it doesn't invent or flatten findings
- Build an opportunity solution tree under one outcome and choose a target opportunity
- List the assumptions behind a solution and pick the riskiest one to test
The everyday picture
At the end of the month you have a shoebox of 200 receipts and a feeling that you spent too much. Reading them one by one gets you nowhere. So you sort them into piles: food, transport, phone, family. Inside transport you notice a smaller pile: taxis on rainy mornings. Now you have something you can act on. You could leave earlier, buy a good umbrella, or share rides with a colleague, and you can try the cheapest idea next week to see whether it helps.
That is this whole lesson in miniature. The shoebox is your research. The piles are opportunities. "Spend less this month" is the outcome. The umbrella and the shared ride are solutions, and "try it next week" is an assumption test.
Synthesis: from stories to opportunities
Synthesis is the work of turning many separate observations into a few patterns you can act on. Done badly, it is the loudest story from the most recent interview. Done well, it is a short list of needs with evidence anyone can check. It takes four moves.
- Atomise. Break each snapshot, field note and ticket into atomic notes: one observation per note, with a source ID (interview 3, ticket 1182, visit 2) so anyone can trace it back.
- Cluster from the bottom up. Group notes that describe the same need, without starting from categories you made up in advance. Affinity mapping is the usual method: notes on a wall or a digital board, moved around until the groups make sense. Name each group as a need in the user's voice, such as "I don't know what I owe", never "Credit feature".
- Count people, not mentions. One angry shop owner who mentions credit nine times is one person. Count how many different people told a story that fits each cluster, and which segment they belong to.
- Weigh. For each cluster, note how often it happens, what it costs when it does (minutes, money, lost sales) and who it affects. A need felt by 600 phone-ordering shops every week outweighs one felt by 20 shops once a quarter.
Teresa Torres defines an opportunity as an unmet customer need, pain point or desire. Her test for whether something really is an opportunity: is there more than one way to address it? "I don't know what I owe" could be met by a balance screen, a weekly message or the rep's call, so it is an opportunity. "Add a balance screen" has only one answer, so it is a solution in disguise.
Quick check
Which of these is an opportunity rather than a solution?
The third describes a need that many different solutions could meet. The others are specific solutions. Translation, for example, is one answer to a different opportunity: "I can't find products by the names I use".