AI agent team guide

AI Win/Loss Analysis: Build an Evidence-Led Learning Loop

A reconciled win/loss evidence register with one testable operating change

By Finta Editorial TeamPublished November 22, 2026Verified November 22, 2026Last reviewed November 22, 2026

A lilac drop-shaped coworker beside a clear two-faceted prism at an emerald river bend.

Explore Sales team guides

AI win/loss analysis should explain what the evidence can support, including where the reasons are unknown. Choose a defined group of completed opportunities, keep wins and losses in the same reporting frame and trace each proposed explanation to notes or customer evidence. Then decide what to investigate or change next.

Win/Loss Analyst owns the learning brief. Sales Pipeline Analyst checks cohort definitions, Customer Voice Researcher examines attributed feedback and Decision Red Team challenges convenient explanations. The goal is a defensible learning decision, not a confident story about motives nobody recorded.

Build your four-agent team

Suggested agent team. Add each template individually, then follow the setup and handoffs in this guide.
Agent templateIts job on the teamBring to its setupReviewable returnTeam

Win/Loss Analyst

Lead agent
Code supported opportunity outcomes and explanations.
A defined cohort, outcome records and attributed notes.
An evidence-coded learning brief with unknown reasons retained.
Add to team

Sales Pipeline Analyst

Supporting agent
Check the denominator and cohort boundaries.
Stage definitions, reporting dates and the included opportunity records.
A reconciliation of eligible wins, losses, exclusions and unresolved outcomes.
Add to team

Customer Voice Researcher

Supporting agent
Separate customer evidence from seller interpretation.
Authorized interviews, feedback and source references.
Attributed themes with contradictory accounts and coverage limits.
Add to team

Decision Red Team

Supporting agent
Test the favored explanation.
The draft brief, evidence ledger and proposed change.
Counterevidence, alternative explanations and a bounded investigation or test.
Add to team

Set up your team in Finta

  1. Add each template individually. Use the Add to team action for each named coworker, review the import setup and complete its Getting Started skill.
  2. Give each role the right context. State the shared objective, then supply the authorized documents, notes or exports that role needs. Do not assume an unconnected system is available.
  3. Review the handoff. Accept or correct one return before giving its evidence and open questions to the next coworker. Importing four templates does not itself run a bulk team.
  4. Make recurrence deliberate. Complete a useful manual pass first. Review any supported routine and enable it only when the scope, inputs and permissions are right.

Prepare the evidence

  • A complete cohort definition, cutoff date and outcome terminology.
  • Won, lost and unresolved records with their actual sources.
  • Permissioned customer feedback and seller notes identified separately.
  • A proposed learning decision and the person accountable for accepting it.

Run the workflow

1. Reconcile the cohort before counting reasons

Sales Pipeline Analyst should explain who is included and why. Do not mix open opportunities with closed losses or compare periods with different definitions without showing the difference. Preserve exclusions and unresolved records.

2. Code what was said, not what would be convenient

Win/Loss Analyst maps explanations to source material. “No decision” is an outcome description, not proof that price caused the loss. Multiple factors can coexist, and a record can have no evidenced reason.

3. Look for contradictory customer accounts

Customer Voice Researcher keeps attributed feedback separate from internal speculation. A vivid interview is not automatically representative. Note how many included opportunities have usable customer evidence and which do not.

4. Choose a small learning action

Decision Red Team tests whether the proposed response addresses the observed pattern or merely the team’s preference. The next step might be better interview coverage or clearer discovery, not an immediate pricing change. Record the hypothesis, owner and review trigger.

A fictional cohort that resists an easy explanation

All counts and records below are invented for teaching. They are not benchmarks or Finta sales results.

Cohort evidenceSupported readingAction
12 completed opportunities: 5 won and 7 lost.The outcome denominator is 12 under this example’s definitions.Keep open opportunities outside this cohort.
Three losses have a seller note mentioning price.Price was noted internally in three records.Do not claim three customers independently named price.
Two customer interviews describe implementation uncertainty.There is attributed evidence for that concern in two accounts.Review the discovery and implementation explanation.
Four losses lack usable reason evidence.Their reasons are unknown.Improve coverage rather than assign them to the most common theme.

The honest brief can be more actionable than a dramatic chart. It identifies which explanation is grounded and which decision needs better evidence.

Copy the team brief

Analyze wins and losses for [defined cohort and cutoff].
Pipeline Analyst: reconcile outcomes, exclusions and the denominator.
Win/Loss Analyst: code each explanation to a source and retain unknown reasons.
Customer Voice Researcher: separate attributed customer statements from internal interpretation.
Decision Red Team: challenge the favored explanation and propose alternatives.
Return cohort coverage, evidence-coded themes, contradictions and one bounded learning action.
Do not infer buyer motives, invent missing interviews or claim causation from a small descriptive cohort.

Review the return

  • The cohort can be reproduced from its definition and cutoff.
  • Customer statements, seller interpretations and unknown reasons remain distinct.
  • The proposed change is proportionate to the evidence and its coverage.

Questions before you start

Can the agents determine why every deal was lost?

Only to the extent the evidence supports an explanation. Missing or conflicting feedback should remain unknown rather than be assigned a plausible motive.

How is this different from a win/loss interview guide?

An interview guide helps collect better evidence. This workflow coordinates cohort reconciliation, coding, counterevidence and the decision that follows the available research.

Sources and next steps

This playbook uses the current published Finta marketplace packages linked in the team table. Open each profile to review its instructions, context, Skills and any routines before importing.