Win/Loss Analysis for Revenue Teams: How to Run It
Leah Clapper
Win/loss analysis explains why deals close or die, using evidence from buyers and from the deal record. Revenue teams use it to fix messaging, qualification, pricing and product gaps.
This guide shows how to build a repeatable program, from sampling to the quarterly readout. All numbers in examples are illustrative.
What is win/loss analysis?
Win/loss analysis is a structured review of closed deals that records the reasons a buyer chose or rejected your offer. It combines two sources: CRM data (stage history, competitors, discount, cycle length) and buyer input (interviews or surveys).
The output is a ranked list of reasons, tied to segments, with an owner for each fix.
A win rate tells you how many deals closed. Win/loss analysis tells you why. The two answer different questions, and the revenue operations dashboard guide covers the rate itself.
Which entities and attributes does a win/loss program track?
A program tracks the deal, the buyer, the competitor, the reason, and the owner. Record the same attributes on every deal so results can be compared.
Entity | Attribute | Example value | Source |
|---|---|---|---|
Deal | Outcome | Won, lost, no decision | CRM stage |
Deal | Segment | Mid-market, enterprise | Account field |
Deal | Cycle length | Days from creation to close | CRM dates |
Buyer | Role | Economic buyer, champion | Contact roles |
Competitor | Name | Named vendor or in-house | Rep field or interview |
Reason | Primary reason | Price, fit, timing, trust | Interview or survey |
Owner | Fix owner | Product, marketing, sales | Program lead |
Add "no decision" as its own outcome. Many lost deals are not lost to a competitor. The buyer did nothing, and the reasons for that differ from the reasons for a competitive loss.
Why do CRM loss reasons fail on their own?
CRM loss reasons are weak evidence because reps select them from a dropdown, often after the fact and often with the least uncomfortable option. "Price" is a common default because it blames no one. Buyers often cite fit, risk or internal priorities when asked directly.
Use CRM data for the facts it holds well: stage history, discount, competitor field, cycle length. Use buyer input for the reasons. Where the two disagree, report the disagreement. It is itself a finding.
How do you choose which deals to review?
Review a sample that covers every outcome and segment, not only the biggest deals. A workable rule is to review all losses and no-decisions above a deal-size floor, plus a matching number of wins from the same segments.
Set a deal-size floor that keeps volume manageable.
Include every outcome: won, lost to competitor, lost to no decision.
Include each segment you sell into.
Match the number of wins and losses so comparisons are fair.
Exclude deals still in dispute or under contract negotiation.
Write the sampling rule down. A rule applied the same way every quarter lets you compare quarters.
How do you collect buyer input?
Collect buyer input through short interviews with someone who was close to the decision, held by a person who did not sell the deal. Buyers are more candid with a neutral party.
Surveys work when interviews are not possible, but they get shorter, more polite answers.
Keep the interview to five topics:
What problem were you trying to solve?
Who else did you consider, and why?
What mattered most in the final decision?
Where did we help or hurt our case?
What would have changed your decision?
Record the answers in the same fields for every deal. Free-text notes alone cannot be ranked.
How do you code reasons so they can be ranked?
Code each interview into a primary reason and up to two secondary reasons from a fixed list. A fixed list of 8 to 12 reasons keeps coding consistent across reviewers.
Reason group | Example codes |
|---|---|
Product | Missing capability, usability, integration gap |
Commercial | Price, packaging, contract terms |
Sales process | Slow response, weak discovery, no multi-threading |
Positioning | Unclear value, wrong persona, weak proof |
Buyer-side | No budget, no decision, internal priority change |
Have two people code the first ten interviews and compare. Where they differ, tighten the definitions. After that, one reviewer can code with a spot check.
How do you turn findings into action?
Turn each ranked reason into a fix with an owner, a due date and a measure. A finding without an owner is a report, not a program.
Finding (illustrative) | Owner | Fix | Measure |
|---|---|---|---|
Buyers could not tell us from a competitor | Marketing | Rewrite comparison page | Share of lost deals citing positioning |
Champion left mid-cycle | Sales | Require a second contact by stage 3 | Share of deals with two or more contacts |
Integration gap named in many losses | Product | Review the roadmap item | Losses citing the gap |
The buying committee mapping guide covers the second contact. The deal risk detection guide covers spotting the same signals while a deal is still open.
How often should you run it?
Run a light review monthly and a full readout quarterly. The monthly review keeps interviews flowing. The quarterly readout needs enough closed deals per segment to see a pattern. If a segment has only a handful of deals, report counts, not percentages.
Share the readout with sales, marketing, product and finance together. Each team owns different fixes, and a shared meeting prevents the findings from reaching only one of them.
Where can agents help?
Agents can assemble the CRM facts for each deal, draft interview invitations and summarize call notes. A person should still hold the interview, code the reasons and approve any change.
Rox agents operate within user-defined parameters and their outputs should be reviewed by authorized staff. See the human in the loop guide for review design.
What mistakes weaken a win/loss program?
Only reviewing losses. Wins show what to repeat.
Letting the rep run the interview. Buyers filter their answers.
Using free-text with no codes. Nothing can be ranked.
Skipping no-decision deals. They are often the largest group.
No owner for fixes. Findings stall.
Changing the sampling rule every quarter. Trends become unreadable.
Conclusion
Win/loss analysis works when it is a routine, not a project. A fixed sampling rule, a fixed list of reason codes and a named owner for each fix turn scattered opinions into a ranked list the whole revenue team can act on.
The weakest link is usually the source of the reasons. CRM dropdowns record what reps chose to log. Buyer interviews record what buyers say they decided on. Use both, and treat disagreement between them as a finding.
Start small. Review the losses and no-decisions above a deal-size floor from the last quarter, code them against a short list, and present the top three reasons with owners. Add wins and more segments once the loop runs.
Results vary with data quality, CRM configuration, sample size and market conditions. Rox does not guarantee specific revenue outcomes.
Frequently asked questions
What is the difference between win rate and win/loss analysis?
Win rate is a percentage of closed deals won. Win/loss analysis explains the reasons behind wins and losses.
How many interviews do I need?
There is no fixed number. Start with the deals you can complete each month and report counts honestly when a segment is small.
Should sales or an outside party run interviews?
A person who did not sell the deal. Buyers are more candid with a neutral party.
Can I use CRM loss reasons alone?
Use them for facts such as competitor and discount. Treat the reason field as a hypothesis to check with the buyer.
What if buyers do not respond?
Keep the survey to a few questions and offer a short call. Report the response rate with the results.
Similar Articles
We build with the best to make sure we exceed the highest standards and deliver real value.
