Win loss analysis

Win/Loss Analysis: The Diagnostic That Improves Everything Downstream

This skill structures a win/loss analysis program that actually changes how the team goes to market. Most companies do some version of win/loss analysis. Almost none do it in a way that produces findings specific enough to act on.

The failure mode is familiar: CRM closed/lost reason fields that contain nothing useful ("not the right time," "went with competitor," "budget"), post-loss debrief conversations with the sales rep that produce the rep's perspective rather than the prospect's, and quarterly summaries that confirm what leadership already believed rather than revealing something that changes a decision.

This skill encodes the practices that produce findings specific enough to change ICP definition, adjust messaging, inform pricing, and redirect channel investment. The analysis is only useful if it feeds back into something that changes.

---

HOW TO SET UP THIS SKILL

Provide:

• What data you currently have on lost deals: CRM closed/lost reasons, rep debrief notes, anything else
• Whether you have access to lost prospects for direct interviews
• The time period you want to analyze
• The primary question you are trying to answer: is this about a specific competitor, a specific market segment, a pricing problem, or general pattern identification?
• What decisions you want to be able to make differently as a result

---

Why CRM Closed/Lost Data Is Almost Always Wrong

Before building any analysis on CRM closed/lost data, understand its limitations. CRM reasons are entered by sales reps under time pressure, often days or weeks after the deal closed, based on whatever the rep believes happened rather than what the prospect said. Four structural problems make this data unreliable:

Reps rationalize losses. A deal lost because the rep did not follow up fast enough, or because the demo was poorly matched to the use case, will be logged as "not the right time" or "lost to competitor." The rep is not lying. They genuinely believe their interpretation. But the root cause is invisible in the data.

Categories are too broad to be useful. "Lost to competitor" tells you nothing about why the prospect chose the competitor. "Price" tells you nothing about whether the price was objectively too high, whether the value was not communicated, or whether the discount was offered too late. Every category in a standard CRM picklist is several real reasons collapsed into one useless label.

Reps only know what the prospect told them. Prospects do not always give honest feedback during a sales process. They tell reps what is comfortable rather than what is true. "We decided to go in a different direction" is easier to say than "your champion left the company and we lost momentum" or "your competitor's CSM flew out to meet with us and yours never called back."

Selection bias on who logs what. Reps who lose deals they expected to win log detailed notes. Reps who lose deals they expected to lose often log quickly and move on. The CRM overrepresents competitive losses on deals where the rep felt invested and underrepresents losses on deals where the rep was never confident.

None of this means CRM data is useless. It means it must be treated as one signal among several, not as ground truth.

---

The Three Data Sources That Produce Reliable Findings

Source 1: Direct prospect interviews (most valuable, hardest to get)

A 20-minute conversation with a decision-maker at a company that evaluated your product and chose not to buy produces more insight than 100 closed/lost CRM records. Prospects are often willing to share their real reasoning once the sales process is over and there is no commercial pressure in the conversation.

Who conducts the interview matters. The prospect will be more honest with a neutral party (a RevOps leader, a product marketer, or an outside researcher) than with the sales rep who lost the deal.

The five questions that produce the most useful findings:

"Walk me through your evaluation process. What triggered the search, who was involved, and how did you approach the decision?"

"What were the two or three things that mattered most to your team when you were making this decision?"

"What was your impression of [your company] during the evaluation? What stood out, positively or negatively?"

"What ultimately drove the final decision?"

"Is there anything we could have done differently that might have changed the outcome?"

Do not defend the product. Do not correct misconceptions during the interview. The goal is to understand what the prospect believed, not to re-litigate the sale.

Source 2: Rep debrief structured correctly

A rep debrief is most useful when it separates the rep's interpretation from what the prospect actually said. These questions produce evidence:

"What did the prospect say, specifically, when they told you they were going with someone else?"

"At what point in the process did you first sense the deal was at risk? What happened that week?"

"Who was the main decision-maker and who else was involved in the final decision? Did you have access to all of them?"

"If you were going to run this deal again knowing what you know now, what would you do differently?"

"What did the competitor do that we did not do?"

Source 3: Pattern analysis across CRM data with corrected categories

Once you have enough interview and debrief data to understand the real reasons, go back and re-code the CRM records. Apply the real categories you have discovered to the full dataset.

---

What to Analyze and What to Look For

The competitive loss pattern: When you lose to a specific competitor repeatedly, the question is not "why is their product better?" The question is "at what point in the evaluation did the outcome tip, and what caused it to tip?"

The internal political loss pattern: Many deals are not lost to competitors. They are lost because the champion left, the budget was reallocated, the project was deprioritized, or the buying committee could not reach alignment.

The messaging gap pattern: When prospects consistently name a capability as the reason they chose the competitor, and that capability exists in your product but was not communicated effectively, you have a messaging gap, not a product gap. The fix is different.

---

Feeding Findings Back Into GTM Decisions

Findings that should change ICP: You are consistently losing in a specific vertical, company size band, or organizational structure.

Findings that should change messaging: Prospects consistently misunderstand a key capability, or consistently name a benefit you provide but did not lead with.

Findings that should change the sales process: You are losing deals at a consistent stage. Investigate what happens at that stage.

Findings that should change pricing or contract structure: You are consistently losing deals where the prospect cited total cost of ownership, implementation timeline, or contract flexibility.

Findings that should change channel investment: You are winning in accounts sourced through a specific channel at a higher rate than accounts sourced through others.

---

Running Win/Loss on Wins Too

Win/loss programs that only analyze losses miss half the picture. The most useful question from a win interview: "At what point did you decide you were going to go with us, and what caused that?"

---

Output Rules

• Be direct about the unreliability of CRM closed/lost data. Do not treat it as ground truth. Flag its limitations before building analysis on it.
• The findings-to-decisions map is required. Win/loss analysis without a mapped decision is a report, not a diagnostic.
• Distinguish product gaps from messaging gaps explicitly. The fix for each is different.
• If the user does not have access to lost prospects for interviews, say so directly and adjust the plan.
• No em dashes. Use commas or periods.