Reviewing why a sale fell through: a table for signals and checks
Use calls and correspondence to analyze why a prospect declined. Separate the customer’s words, deal stage, events, interpretation, category and follow-up test.
A final remark may just be a polite way to end the call
A customer may say “we’ll think about it” after an implementation question goes unanswered. If your CRM stores only that remark, the team may fix the wrong thing. Trace the events from initial interest to the end of the discussion.
Transcripts make conversations comparable. A category remains a hypothesis until words, events or further checks support it. Contact the customer again only if you have a lawful basis and an agreed process.
Trace the decline through the sequence of events
Identify where the deal stopped
Record the last agreed step and what should have happened next. Distinguish an explicit refusal, no response and a postponed decision.
Find earlier signals
Review questions, doubts, delays, changes in participants and promises that weren’t kept. Keep timestamps alongside the final label.
Separate facts from interpretation
Fact: the customer asked about moving data, and the recording contains no answer. Interpretation: implementation risk may still have concerned them. Keep those in separate fields.
Choose an action that can test the hypothesis
Match the action to the signal: check demo materials, coach the team on a question, change qualification or collect more interviews. Compare outcomes before promising an improvement.
A lost deal recorded as “too expensive”
This is a fictional training example. The names, circumstances and outcomes are included solely to explain the method.
The price is above our budget, but the main question is who will move our archive and how long staff will have to work in two systems.
The customer raised both price and implementation. If the seller answered only with a discount, operational risk may still have been the sticking point. Record both factors until there’s evidence to isolate one.
To test it, look for a clear migration plan, compare similar won and lost deals and check whether migration came up before price. The scenario is fictional.
| Field | Example entry | Status |
|---|---|---|
| Direct signal | Question about moving the archive | Confirmed in the recording |
| Additional factor | Price above budget expectations | Stated by the customer |
| Interpretation | Implementation risk remained | Hypothesis |
| Checks | Compare responses and deal outcomes | Next action |
Lost deal review table
Use consistent fields so you can compare cases without changing the categories to fit each one.
Deal / anonymized ID: Stage and stop date: Last agreed step: Customer’s exact words: Questions and signals before the decline: Seller’s responses: Unfulfilled promises: Factors mentioned: Analyst’s interpretation: Confidence level and basis: Timestamps / sources: Category from the reference list: What to verify: Action and owner: Review date:
Check the quality of the categories
- Explicit refusals are distinct from silence and delays.
- Direct quotes are separate from interpretation.
- The category is based on events.
- The action tests a hypothesis without promising growth.
A closer look
How does a lost deal differ from churn?
A lost deal is a prospect that never closed. Churn is when usage stops or shrinks after purchase, so it requires different context and data.
Can we count reasons?
Yes, if categories and coding rules are stable and cases are comparable. Check the categories first, and keep assumptions separate from confirmed causes in the count.