Atul Singh logoAtul Singh
← Back to prompts
RevOpsRevOps Analystintermediate 45 min saved

CRM Pipeline Integrity Auditor

Audit CRM deals against stage criteria to identify stale or prematurely advanced opportunities with evidence-based reporting.

Customise & copy

Add the context you have. The prompt updates as you work.

Provide the explicit entry and exit criteria for each sales stage.

List opportunity names, current stages, and age in stage.

Paste activity logs, emails, or notes detailing actual deal history.

Specify the segment or region being audited.

Select the level of detail required in the audit report.

Live prompt

5 fields
You are a rigorous CRM pipeline auditor. Evaluate the provided CRM opportunities strictly against the defined sales stage criteria to identify misaligned, prematurely advanced, or stale deals. Base your findings exclusively on the provided evidence without making assumptions about offline activities or missing data.

Sales Stage Definitions:
Stage 1: Qualification - Discovery call completed, budget confirmed. Stage 2: Evaluation - Technical demo delivered, stakeholders identified. Stage 3: Proposal - Formal quote sent, legal review initiated.

Opportunity Records:
Opp 101 - Acme Corp: Current Stage: Proposal, Age in Stage: 45 days. Opp 102 - Beta Inc: Current Stage: Evaluation, Age in Stage: 12 days.

Activity Logs, Emails, and Notes:
Opp 101: Last email sent 45 days ago asking for meeting. No response, no quote sent, no legal review. Opp 102: Demo completed yesterday with CTO, budget confirmed on call.

Target Pipeline Segment:
Enterprise North America

Audit Depth:
Standard (Flag exceptions only)

Instructions:
1. Audit each opportunity by comparing its current stage against the explicit entry and exit criteria provided in the stage definitions.
2. Classify each opportunity strictly as 'Correctly Placed', 'Prematurely Advanced', or 'Stale'.
3. For any flagged exception (prematurely advanced or stale), cite explicit evidence from the activity logs mapped directly against the required criteria. Do not infer missing evidence or assume unrecorded offline progress.
4. Output a structured audit report listing every evaluated opportunity, its classification, and corresponding evidence citations.

Proof before you use it

A real example

Tested on gemini-3.1-flash-lite on 2026-09-06

{
  "audit_depth": "Comprehensive (Detail all checks per deal)",
  "pipeline_segment": "Healthcare SMB EMEA",
  "activity_evidence": "Opp 889: Trial invitation sent 30 days ago, but login metrics show 0 platform logins to date. Opp 890: Calendar invite accepted for tomorrow at 2 PM.",
  "stage_definitions": "Stage 1: Lead Inbound - Form filled. Stage 2: Demo Booked - Calendar invite accepted. Stage 3: Trial Active - Software login provisioned.",
  "opportunity_records": "Opp 889 - Hospital Group A: Current Stage: Trial Active, Age in Stage: 30 days. Opp 890 - Clinic B: Current Stage: Demo Booked, Age in Stage: 3 days."
}

A small ritual that works

How to use it

  1. 01Paste your specific sales stage entry/exit criteria into the definitions field.
  2. 02Input your current CRM opportunity list and corresponding activity logs.
  3. 03Select your desired audit depth and run the prompt to generate the report.
  4. 04Review the flagged exceptions to clean up your pipeline data.

Why it works

The prompt enforces rigorous, strict constraints on pipeline evaluation by forcing the LLM to map activity evidence directly against explicit stage entry and exit criteria. It eliminates subjective guessing by explicitly forbidding the inference of offline activities or missing data.

Where it fails

LLMs may conflate the lack of activity with failure to meet a specific stage criterion (e.g., treating zero logins as proof a login was not provisioned), or accidentally assign multiple classifications to a single opportunity instead of choosing one strict label.

Customisation tips

Modify the stage definitions variable with your organization's exact entrance and exit requirements, and adjust the audit depth setting depending on whether you need a high-level exception report or a comprehensive deal-by-deal breakdown.

Watch-outs

  • Ensure the model strictly adheres to a single classification per opportunity ('Correctly Placed', 'Prematurely Advanced', or 'Stale').
  • Prevent the model from hallucinating or adding unsupported metadata such as external audit dates.
  • Check that the model does not recommend unrequested deal dispositions like moving deals to 'Closed-Lost' or changing pipeline stages.
  • Ensure the model strictly cites recorded evidence without assuming unrecorded offline progress.

Works in

SalesforceHubSpotPipedriveChatGPT

Was this useful?

Submit your variation

If your version helps, it may be published with your name and a link back to you.

Related prompts

RevOps

CRM Revenue Data Integrity Auditor

Audit CRM exports for data quality issues, operational risks, and actionable cleanup rules to ensure reporting and automation accuracy.

Want to make AI useful across your team? Explore Atul's training.