As of 13 August 2026, AI can analyse why your deals are lost.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 1 hour until you can act on the result.

   How to actually do it

   1. Open your CRM and export lost deals for a defined period, including stage reached, stated loss reason, segment, salesperson, competitor if known, dates and activity history.
   2. Export a comparable sample of won deals and keep the same fields so the model can compare outcomes rather than inspect losses in isolation.
   3. Gather relevant call transcripts, meeting notes and email excerpts from the deal records, then remove names, contact details and commercially sensitive details that are not needed.
   4. Create a short context note defining your product, target buyer, sales stages, segments and any changes to pricing, packaging or process during the period.
   5. Paste the context, anonymised data and the copyable prompt into a chatbot, then ask it to analyse the evidence and identify data gaps instead of filling them.
   6. Check every theme against the cited deal IDs, excerpts and CRM records, correcting classifications where the model has confused a stated reason with a confirmed cause.
   7. Ask a sales manager or colleague who knows the accounts to challenge the findings, then agree a short list of actions and tests rather than treating the analysis as proof of buyer intent.

   Prompt

   Analyse the lost deals in the data below and explain why they were lost.
   
   Context:
   - Business and market: [brief description]
   - Product or service: [description]
   - Typical buyer: [job title, organisation type and relevant segment]
   - Sales stages: [stage names and definitions]
   - Period covered: [date range]
   
   Data:
   [Paste an anonymised table or export containing deal ID, segment, value band, stage reached, salesperson, competitor if known, stated loss reason, dates, activity summary, relevant email excerpts and call transcript excerpts. Include comparable won deals where available.]
   
   Instructions:
   1. Use only evidence in the supplied data. Do not invent causes, buyer motives, competitors, figures or missing events.
   2. Separate direct evidence, reasonable inference and unknowns.
   3. Compare lost deals with won deals where the data allows it, and state when the comparison is not reliable.
   4. Group losses into clear themes, showing the supporting deal IDs or excerpts for each theme.
   5. Distinguish causes within our control from external conditions and from recording or process problems.
   6. Flag contradictory evidence, small or biased samples, duplicate records and inconsistent loss reasons.
   7. Do not treat a salesperson's stated loss reason as confirmed buyer truth unless the records support it.
   8. Produce this structure: executive summary, evidence table, recurring themes, deal-by-deal findings, confidence and data gaps, actions to test, and questions for the sales team or customers.
   9. Recommend no more than five actions. For each action, state the evidence, the owner, what should change and how the team could test whether it helped.
   10. Keep customer and personal data anonymised, and quote only the minimum text needed to support each finding.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.