Product marketing
Win/loss analysis without a research team
You do not need a research function to run win/loss. You need five questions, a rule about who asks them, and somewhere the answers cannot rot.
SteveWin/loss has a reputation as a research programme, which is why most teams never start one. It is five questions, asked by someone who did not work the deal, within two weeks of the decision.
Key takeaways
- Five questions is a programme. Thirty is a survey nobody finishes.
- The person who lost the deal cannot run the interview.
- Ask within two weeks, before the buyer tidies the story.
- Close-reason fields record the rep's belief, which is the thing you are checking.
- Feed the findings back into the battlecards, or the exercise is journalism.
8–12
Interviews per quarter
20 min
Time per interview
2 weeks
Ask within
5
Questions
The five questions
When did you decide?
Almost always earlier than the rep thinks. The answer reframes everything that follows.
Who else was in the room?
Names the real committee, which the CRM usually does not have.
What did you compare us against?
Including "nothing" and "a spreadsheet", which are two of the most common answers and neither appears in a competitor field.
What nearly changed your mind?
The single most useful question. It surfaces the argument that almost worked.
What would we have had to do?
Ask it last. Asked first, it invites a feature request instead of a story.
Who asks
Not the rep. This is the part teams get wrong, and it is not about honesty — a buyer will simply not tell the person they rejected why they rejected them, and a rep cannot hear "your pricing page confused us" without defending the pricing page.
Where the answers go
A findings deck read once is journalism. The output has to land somewhere a rep already looks, which in practice means the battlecards.
| Finding | Deck | Battlecard |
|---|---|---|
| "We compared you to a spreadsheet" | Slide 6 | New card: the do-nothing objection |
| "Their security review blocked us" | Slide 9 | Proof line, dated |
| "We decided in week one" | Slide 3 | Discovery question, not a card |
The quarterly rhythm
- Pull closed-won and closed-lost from the last 90 days
- Pick eight, half won — a loss-only programme reads as an autopsy
- Send the two-sentence ask from someone not on the deal
- Twenty minutes, five questions, recorded with consent
- One page of patterns, and the battlecard edits they imply
What this catches
- Which competitor was actually in the deal
- When the decision was really made
- The argument that nearly worked
- Objections nobody on the team has heard yet
What it will not
- Market sizing
- Anything statistically significant
- Why a segment you never sold into ignores you
Frequently asked
How many interviews make win/loss analysis useful?
Eight to twelve per quarter is enough to see a pattern. Below five you are collecting anecdotes; above twenty you are running a research function, which is a different job.
Should the rep who lost the deal run the interview?
No. The buyer will not tell the person they rejected why they rejected them. Someone one step removed gets a materially different answer.
How soon after the decision should you ask?
Within two weeks. After a month the buyer has rewritten the story into something tidier than what actually happened.
Can CRM close-reason fields replace interviews?
No. Close-reason fields record what the rep believes, which is the thing win/loss exists to check.
Keep reading
All posts- September 2, 2026A three-line battlecard format for live sales callsUse one concise card per objection: a one-sentence answer, one dated proof point and the likely rebuttal. Then test the format with sellers.
- October 7, 2026What ChatGPT, Claude, Gemini and Perplexity say about you vs your top competitor: a ten-prompt auditRun ten fixed prompts three times each on four AI engines, memory off and region set, then report how often you are named and which pages shaped the answer.
- October 7, 2026Can Claude or ChatGPT replace your CI tool? A ten-task test you can run in an afternoonRun ten competitive intelligence jobs on Claude and ChatGPT, score each on correct, sourced, fresh and unattended, and see where an assistant is enough.
- October 7, 2026Prove CI impact without attribution theatre: three defensible metrics and a worked exampleSkip the ROI multiple. Report competitive win rate on comparable deals, card use in won versus lost deals, and response time, with a dummy worked example.
See what Steve finds in your market