Win-loss analysis means reading your won and lost deals to understand what made the difference. Most teams run it through buyer interviews. There is a faster source, and you already own it: everything your deals recorded, email by email and meeting by meeting, up to the day they closed.
This guide walks through a simple way to read that source by hand, on a spreadsheet: the ten-loss method. The worksheet is at the bottom of the page.
The short version
- The loss reason in your CRM tells the story from the seller's side. What the buyer did in the weeks before tells a different one, and that is the story worth reading.
- Ten losses are enough to start, read next to ten wins. You are looking for what repeats from one deal to the next.
- Read every deal up to the day before it closed. The signature day produces patterns that look perfect in hindsight and tell you nothing about an open deal.
- Every question has three answers: yes, no, and "can't say yet".
Why the CRM loss reason falls short
"Budget freeze", "bad timing", "went with a competitor", "not a priority this year". The loss reason gets typed after the fact, by the person who lost the deal, often on the day the line has to be closed. It records how the team explains the loss.
The buyer's moves are dated, and nobody rewrites them later: the last reply, the last meeting they proposed, the last colleague they brought in, the last document they sent. Putting the two side by side is usually the most useful part of the exercise. A "budget freeze" logged in December reads differently when the buyer's last email is from September.
"Ten years of SaaS sales, and I still got fooled by an 'it's in the bag' in week eleven. Nudge started there." Ridha Mami, founder of Nudge
The ten-loss method
- Take your last ten losses, and ten wins from the same period. Ten losses are enough to start. The wins are your control: a move that keeps showing up in the losses matters most when the wins lack it.
- For each deal, write down the buyer's last move and its date. Count only what the buyer did: a reply, a meeting they proposed, a colleague they brought in, a document they sent. A follow-up that got no answer is still a seller's move. Add the stage where the deal stalled, and the reason logged in the CRM.
- Read each deal up to the day before it closed. The signature email, the "we've decided" email, the recap and the debrief written afterwards are the outcome of the deal. Leave them in and a line like "we'll sign", written on signature day, becomes a perfect pattern: perfect in hindsight, and useless on a deal that is still open.
- Answer every question with yes, no, or "can't say yet". A deal that never got a proposal can't tell you whether the economic buyer was there before the proposal. Counting it as a "no" makes every young deal look like a loss.
- Compare wins and losses, one question at a time. Look for what the wins had and the losses lacked: who the buyer brought in, and when they put a date on a next step. Look at names as well: the ones that show up in wins before pricing, and stay missing from the losses until it is too late.
The questions to ask of every closed deal
Some of these you can answer from CRM fields and dates in a few minutes. The others need you to reread the emails and meeting notes.
| Question | Where to read it |
|---|---|
| Was the close date pushed twice or more? | CRM history |
| Did the buyer go silent for 21 days or more? | Dates of the exchanges |
| Was the economic buyer engaged before the first proposal? | Participants and emails |
| Was the deal single-threaded on the buyer side? | Participants |
| Did a new stakeholder join within 20 days of the first meeting? | Participants and dates |
| Did the buyer start email threads themselves? | Emails |
| Did the seller follow up three times or more without a reply? | Emails |
| Did the champion go quiet in the last 14 days of the deal? | Dates of the exchanges |
| Did procurement or legal take part? | Participants |
| Did the buyer put a number on the problem themselves? | Text of the exchanges |
| Was the signing process known before pricing? | Text of the exchanges |
| Did the buyer own a dated next step? | Text of the exchanges |
Each question describes something the buyer did. The point is to learn how your buyers buy, so the reps stay out of it.
Buyer interviews or deal reading?
Use both. Interviews tell you why the buyer believes they decided. Reading the deal tells you what they did, and when.
| Buyer interviews | Reading your closed deals | |
|---|---|---|
| What you get | The reason the buyer gives, in their own words | The buyer's moves, with dates |
| What it takes | Buyers willing to talk to you after the decision | Your email, calendar, meeting notes and CRM |
| The limit | A memory rebuilt after the decision | Whatever the deal never recorded, like a call nobody logged |
| Best for | Understanding one important loss in depth | Spotting what repeats across deals |
What win-loss analysis can't prove
A pattern in your history shows what went together in the deals you won. It can't prove that repeating it will win the next one. Treat every pattern as a hypothesis to check against your open deals, and keep the ones that hold.
How sharp the reading gets also depends on what you can see. A deal run mostly by phone leaves little behind: mark it as such, and draw no conclusion from a deal you can only half see.
The worksheet
Download the ten-loss worksheet (CSV, opens in Excel or Google Sheets). It has twenty rows, ten losses and ten wins, and a column for each key question. Fill it in deal by deal, then compare the two blocks.
Nudge runs this reading across all of your closed deals, then watches your open deals for the same moves. To see what that looks like on your own deals, start with a 30-minute review.
FAQ
What is win-loss analysis?
Reading your won and lost deals to understand what made the difference. It can draw on buyer interviews, and on what your deals recorded along the way.
Can you do win-loss analysis without interviewing customers?
Yes. Your closed deals already hold the buyer's moves: replies, meetings, people brought in, next steps agreed and kept. Reading them shows what the buyer did. Interviews add the reasons, and they are worth it for an important loss.
How many deals do you need to start?
Ten losses and ten wins from the same period are enough to see what repeats. More makes the exercise more reliable, and takes longer to read.
How often should you run it?
A full read every quarter works well, ahead of the board review. In between, run each new loss through the same questions.
How is this different from a pipeline review?
A pipeline review looks at open deals. Win-loss analysis looks at closed ones, to learn what separates a buyer who keeps moving from one who drifts off. Each feeds the other: what you learn from closed deals becomes the questions you ask in the pipeline review.