
How to Run a Win Loss Analysis for a Service Business: Step-by-Step Guide
Start a sales win/loss review by comparing completed opportunities with the same questions: what the buyer needed, what happened during the sale, and what evidence supports the recorded outcome. Keep the buyer's stated reason separate from your team's interpretation.
A lost proposal can make you question the price, the offer, or the conversation. Before changing any of them, check what you actually know.
For women business owners and sales teams, this review can be a useful alternative to a meeting that turns into blame or guesswork. The aim is to choose a specific improvement you can test.
Choose a comparable set of opportunities
Start with one service and a defined review period. Include wins as well as losses so you're not drawing conclusions only from conversations that disappointed you.
Decide what belongs in the review. A buyer who declined a proposal is different from an inquiry that never became a sales conversation. A deal still under discussion isn't a completed loss.
Record the outcome using your team's agreed definitions. If the buyer stopped responding, label that as no response or an unknown reason. Don't turn silence into “too expensive.”
Microsoft's opportunity-closing guidance illustrates why the closing record matters: it describes recording details such as close date and description, then reviewing those details to understand wins and losses. You can use the same discipline in the CRM or spreadsheet your team already maintains.
Build a review record before asking AI for patterns
For each opportunity, gather the relevant notes and messages your business is permitted to use. Keep the record short enough that someone else can check it.
Buyer need: What problem or goal did the buyer describe?
Service discussed: What did your team propose, and what was outside the offer?
Outcome: Was it won, declined, deferred, or closed without a confirmed reason?
Buyer's explanation: What did they actually say, and where is it recorded?
Sales actions: Which questions, documents, and next steps were handled or left open?
Interpretation: What does the team think may have affected the decision?
Unknowns: What can't the available record establish?
Use quotes accurately and keep the source with them. If a salesperson remembers a comment but it isn't documented, mark it as their recollection rather than a verified quotation.
I would rather see “reason not confirmed” than a neat category that misrepresents the conversation. The unknown is something you can investigate.
A worked example: timing is not the same as price
Consider this fictional example. Nia owns a consultancy that offers marketing implementation support. A buyer declines after receiving a proposal and writes, “We can't allocate staff to this before our November launch.”
The team had assumed that the proposed fee was the problem. The message supports a concern about staff availability. It doesn't establish that price was acceptable or unacceptable.
The review record could say:
Outcome: buyer declined the current proposal. Stated reason: staff unavailable before the November launch. Evidence: buyer's email. Team interpretation: implementation workload may not have been clear early enough. Unknowns: whether budget also affected the decision, and whether the buyer wants to revisit the project later.
A useful improvement would be to ask about team availability earlier in future discovery conversations and explain the client's responsibilities before preparing the proposal. That is a testable change. The record doesn't support cutting the price or promising that this change will win the next sale.
Use AI to organize evidence, not invent buyer motives
AI can help organize review records and identify recurring questions. Use a business-approved tool and account, with identifying or confidential details removed when they're unnecessary.
Paste this prompt above the prepared records:
Review these completed sales-opportunity records using only the supplied evidence. For each record, separate the outcome, buyer-stated reason, team interpretation, and unknowns. Cite the record label and supporting note for each observation. Group recurring issues only when the records support the grouping. Do not infer motives from silence, invent quotes, treat a deferred or open deal as a loss, or claim that one factor caused an outcome. Return possible patterns for human review and questions we need to investigate. Suggest small process changes as hypotheses to test, not proven solutions.
Check each proposed pattern against the original records. “Three buyers mentioned implementation workload” is a narrower observation than “our delivery model is why we lose deals.” The second statement needs more evidence.
Ask a follow-up question only when it fits the relationship
If a buyer is open to sharing feedback, ask a short question about the decision. Don't turn the request into another attempt to overcome their objection.
For Nia's fictional example, a question for review could be: “Thanks for explaining the staffing constraint. If you're comfortable sharing, was anything about the proposed work or decision process unclear? Your feedback would help us explain future projects more clearly.”
Review whether that message is appropriate before sending it. Respect a request for no further contact, and don't record an answer the buyer hasn't given. A declined proposal doesn't create permission for repeated follow-up.
If the record is missing basic next steps, start with the AI sales follow-up workflow. Better notes make the later review easier to check.
Choose one change and review it again
Give the improvement an owner and decide how you'll observe whether it helps. In Nia's example, the sales lead could add a client-capacity question to discovery notes and check whether future proposals leave workload questions unanswered.
Track the process change alongside outcomes, using the same definitions and review period. Changes in buyer mix, service, timing, or lead source may also affect the result. A small set of opportunities can raise useful questions without proving causation.
Keep a record of what you changed so you can decide whether to continue, adjust, or stop. This guide to repeatable service-business systems explains how to document a process your team can use consistently.
If your sales review keeps ending with opinions and no next action, contact Dani about the question you need to resolve. Bring a few comparable opportunities and the evidence behind their recorded outcomes.
