PromptSharp › Prompt Library › Sales › Win-loss read: what actually correlates with winning here

Win-loss read: what actually correlates with winning here

Everyone has a theory about why deals close. Test the theories against your own closed deals and find which ones the data can support.

The prompt — copy and run it

You are a RevOps analyst running a win-loss analysis. I will paste closed deals with their attributes and outcomes.

Produce:

A) THE CORRELATION READ: each attribute I provided against win rate, with the sample size on each side and the arithmetic from my numbers. Any attribute below a usable sample is labelled "too thin to read" and reported separately — never quietly included in a ranking that implies confidence.

B) THE CONFOUND CHECK: for each apparent driver, the alternative explanation my data cannot rule out — deal size, segment, source, rep, or time period — and the cut that would separate them. Say plainly which findings are associations rather than causes, because in this dataset most of them will be.

C) THE ACTIONABLE SHORTLIST: the findings that survive the confound check, each translated into a behaviour a rep or manager can change next week, plus the loss reasons that are recorded but that my data cannot corroborate.

Inputs: [CLOSED-WON AND CLOSED-LOST DEALS: SEGMENT, SIZE, SOURCE, CYCLE LENGTH, STAGES ENTERED, COMPETITOR, RECORDED LOSS REASON, REP OR TEAM] · [THE THEORIES CURRENTLY BELIEVED] · [ANY MULTI-THREADING OR ACTIVITY DATA]

Rules: Do not invent deals, attributes, or win-rates, and never present a correlation as a cause — label every finding association or cause explicitly, and mark thin samples "too thin to read". Keep confidential CRM, deal, and customer data out of consumer AI tools and follow your employer's AI-use policy. This reads the history; you validate any finding on a later period before changing how the team sells. Verify any pattern this surfaces on a later, unseen period before you change how the team sells.

Why this prompt works

Win-loss decks routinely turn a segment effect into a methodology mandate, and a team then changes its behaviour on a pattern that was really deal size all along. Forcing an explicit confound check on every apparent driver is what separates a finding worth acting on from one worth investigating, and the recorded-loss-reason comparison usually reveals that the CRM field is measuring how reps felt rather than why deals were lost.

Frequently asked

Related Sales prompts