Win-loss analysis is a regular review of closed deals, won and lost, to find out why buyers decided as they did. RevOps or enablement usually runs the program and front-line managers use it. A monthly cycle of five to ten deals takes a few hours of interviews and an hour of review.
This version skips the quarterly deck that reps never see. It is built for the manager, so each finding reaches the deals still open this quarter.
What the job is
The job has two halves. The first is learning what decided each closed deal, from the buyer's side as well as the rep's. The second is using it: finding the open deals that show the same pattern and coaching them before they close the same way.
- 01Pick five to ten closed deals
- 02Rep's view, then the buyer'sInterviewed by someone other than the rep
- 03Code each outcome to evidence
- 04Find the gap in open deals
- 05Coach those deals this week
- 06Count the codes each quarterThe top code becomes the coaching focus
- 01Pick five to ten closed deals
- 02Rep's view, then the buyer'sInterviewed by someone other than the rep
- 03Code each outcome to evidence
- 04Find the gap in open deals
- 05Coach those deals this week
- 06Count the codes each quarterThe top code becomes the coaching focus
The link between the two is evidence. If every outcome is coded to the buyer evidence that was present or missing at each stage, a finding reads as "we lose when the signer never heard the case," and a manager can search the pipeline for deals where that is true today. Our deal risk guide lists the evidence each stage should carry.
The procedure: eight steps
Pick the deals each month. Five to ten closed deals from the last 30 to 60 days: wins, losses and deals lost to no decision, above a size that matters. What good looks like: the same selection rule every month, so the counts compare.
Record the rep's view first. Before any buyer interview, the rep writes one line on why the deal was won or lost and fills an evidence line: did the signer hear the case, did a buyer state a cost, was the date the buyer's, were approval steps mapped. What good looks like: a written view that cannot be revised after the buyer speaks.
Interview the buyer, with someone other than the rep. Buyers are more candid with someone other than the person they turned down, and Klue, which sells win-loss software, argues the sales team should not run the interviews at all. A manager from another team, RevOps or an outside firm can run a 20 to 30 minute call. What good looks like: an interview within a few weeks of the decision, using the questions below.
Compare the two views. This is where most of the learning sits. Anova Consulting matched reps' stated loss reasons with what buyers told an independent interviewer: reps had the full and accurate picture 40% of the time, part of it 28% of the time, and were completely wrong 32% of the time. What good looks like: each gap between the rep's line and the buyer's account written down.
Code each outcome to evidence. Use fixed codes for what was present or missing, and add the buyer's stated reason as a second code: signer never heard the case, no cost stated by a buyer, approval steps unmapped, alternative unaddressed, date was the rep's, plus product fit or price where the buyer named them. Keep the codes the same every month so the counts mean something. What good looks like: two reviewers would code the same deal the same way.
Find the same gap in open deals. For the top code this month, search the current pipeline for deals where that evidence is missing today. What good looks like: a named list of open deals, by rep, with the gap stated.
Coach those deals this week. Each deal on the list gets a planned buyer conversation and a date. The largest get a 20-minute coaching call where the rep talks the deal through and leaves with the opening for that conversation, said out loud. What good looks like: no deal on the list without a booked conversation.
Count the codes each quarter. Report the counts by code and by rep to sales leadership and enablement. The top code becomes next quarter's coaching focus, and it feeds the questions in the forecast call. What good looks like: the top loss code falls the quarter after you coach it.
Win-loss interview questions
Ask about the process before the result. Buyers find "who was in the final meeting" easier to answer than "why did you choose them," and the answer often shows the reason.
- Who was involved in the decision, and who had the final say?
- What problem were you solving, and what did it cost you to leave it alone?
- What else did you consider, including doing nothing?
- What set the timing?
- What did the approval process involve after the decision was made?
- What nearly changed the outcome?
- What would you tell the rep to do differently?
A coding sheet for the monthly review
One row per code. The last column is what turns a code into a question managers can ask of an open deal.
| Code | Counts as present when | The deal review question |
|---|---|---|
| Signer heard the case | The person who signs was on a call where the cost and the fix were discussed | "Has the signer been on a call, and what did they say about the cost?" |
| Buyer stated a cost | A named buyer put a number on the problem, or named something they stopped doing | "Who said what this costs, and in which call?" |
| Approval steps mapped | A buyer walked the rep through the steps after a yes, in order | "What happens after they say yes, and who told us?" |
| Alternative addressed | The rep knows what else the buyer weighed, including doing nothing | "What are they comparing us with, in their words?" |
| Date was the buyer's | The close date ties to an event on the buyer's side | "What happens on their side if this slips a month?" |
| Buyer's stated reason | Product fit, price or timing, in the buyer's words from the interview | Second code only; record it, do not coach from it alone |
When a lost deal is missing several, mark them all and name the first gap in stage order as its primary code. The quarterly count uses the primary code, so one bad deal cannot fill four rows.
A worked example
This example is illustrative. A team reviews eight closed deals in a month: three wins, four losses and one no-decision. In the rep write-ups, three losses are put down to price. In the buyer interviews, two of those three buyers say the CFO was never convinced the problem was worth solving this year, and that the price only became an issue once that doubt was there.
Coded to evidence, five of the eight deals (four losses and the no-decision) had no point where the signer heard the case. All three wins did.
The manager searches the open pipeline for deals past discovery where the signer has not been on a call. There are eleven across six reps. Each gets a planned conversation with the champion about bringing the signer in, with a date. The four largest get coaching calls that week. At quarter end, the manager counts again: the share of losses coded "signer never heard the case" has fallen, and two of the eleven deals closed with the signer engaged.
Where it goes wrong
- Letting the rep's reason stand. If the analysis takes the CRM loss reason at face value, it will mostly confirm what reps already believe. Anova's data is the reason to ask the buyer.
- Codes that change every quarter. Shifting categories hide whether a problem is getting better or worse. Fix the codes and live with them for a year.
- Findings that stay general. "Improve discovery" tells a manager nothing to do on Monday. A code tied to missing evidence does.
- A deck in place of deals. Most programs stall between the insight and the rep. If no open deal gets coached the week a finding lands, the program is research.
- Only studying losses. Wins show what evidence was present when deals closed. Without them, the team learns what fails and has nothing to compare it with.
Doing it for every rep
Steps 6 and 7 are the expensive ones: checking every open deal for the gap and coaching each one is more than a manager with eight reps fits into a week. The rep brings each flagged deal to Caty on a call and leaves with the conversation that fills the gap, planned in the team's own method, and the manager sees the pattern across the team through Bryce. The how to improve forecast accuracy guide runs the same loop on slipped commits. If your last win-loss review ended as a deck, talk to us.
FAQ
Who should run win-loss analysis?
RevOps, enablement or product marketing usually owns the program, and someone outside the deal team should run the buyer interviews. Front-line managers should own what happens next: finding open deals with the same gap and coaching them. Reps contribute their own view before the interview.
How many deals should a win-loss analysis include?
Five to ten closed deals a month is enough for a team to see patterns within a quarter, as long as the mix includes wins, losses and no-decisions. Larger programs interview more. A small monthly cycle has one advantage over a big annual study: its findings arrive while the deals they apply to are still open.
How soon after a decision should you interview the buyer?
Within a few weeks of the decision, while the buyer still remembers who was involved and what nearly changed the outcome. Longer gaps produce vaguer answers, and the buyer may have moved on from the role.
What is the difference between win-loss analysis and a deal review?
A deal review looks at one open deal to decide what to do next. Win-loss analysis looks across closed deals to find patterns. They connect when a win-loss finding becomes a question in every deal review, such as whether the signer has heard the case.
Can you do win-loss analysis without buyer interviews?
You can code deals from CRM data, call recordings and rep write-ups, and that is better than nothing. It will lean toward what the rep believed. Anova's data found reps fully right about why they lost only 40% of the time, so add buyer interviews when you can.