Win and loss analysis is a structured interview with someone who recently went through the decision to buy your product or not buy it. Pragmatic Institute puts it in the Market category alongside market problems, because the buying process is one of the few places where a person's stated priorities were tested against their money.
Why it is a separate activity from a sales debrief
The salesperson was a participant. Buyers manage that relationship, which means the reasons they give at the end of a deal are chosen for their effect. Price and a missing feature are the two most common, and both are frequently a polite summary of something else, such as a decision made before the evaluation started or an internal champion who lost an argument.
An interview run by someone with no stake in the deal, some weeks after it closed, gets past most of that. The interviewer can also ask the question no salesperson can ask, which is what would have had to be true for the answer to be different.
What to ask about
What triggered the evaluation. Something changed. Finding out what tells you which market problem became urgent and when.
Who was involved and what each of them wanted. This is where buyer personas come from, and it usually reveals a participant nobody in the pipeline knew about.
What alternatives were considered. Including doing nothing, building it internally and using a spreadsheet. The set of alternatives defines the category the buyer put you in, which is a positioning input.
How the decision was actually made. Criteria, sequence, who had a veto. The buying process, in the framework's words.
What nearly changed the outcome. The most useful question in the interview, and the one that does not appear on any survey.
Both outcomes are needed
Losses show what the market found unconvincing and which competitors keep appearing. Wins show which of your claims did the work and what almost derailed it. Running only losses produces a roadmap of competitor features. Running both produces positioning.
Making it a habit
The activity fails when it is a project. A steady cadence, a small number of interviews per month across both outcomes, feeding the same market problems table, is what makes patterns visible. One round of ten interviews produces anecdotes. Ten rounds of ten produce a trend you can act on.