Answer first
Start with labels, not a live pipeline.
A safe first use is a local packet containing only fictional or organization-approved, minimum-necessary labels: source, owner, time/version, stated purpose, type, and an honest status of supplied, unknown, or conflicting. AI can format these labels and draft neutral questions. Authorized people retain CRM, forecasting, pricing, claim, employment, customer, and communication decisions.
Mission outcome
What you make in one focused session.
Source register
Supplied metric, claim, and coaching-reference labels bound to an owner and time/version.
Unknown log
Gaps and conflicts that remain visible instead of becoming invented analysis.
Review handoff
Neutral questions and an authority map ending only at needs_review.
Field map
Six bounded steps
Choose a fictional or approved label-only scope.
Name the purpose and accountable owner. Exclude CRM exports, account details, contact data, rep notes, calls, emails, compensation, pricing, contracts, and personally identifiable information.
Register supplied labels without treating them as facts.
Capture only source label, owner, time/version, purpose, type, and status. Do not ask AI to infer intent, performance, likelihood, or missing values.
Use AI as an organizer, not a forecaster or scorer.
AI may format fields, surface duplicates, and group blanks. It must not retrieve data, score a lead or person, rank opportunities, predict revenue, or recommend a quota, territory, price, discount, or commitment.
Turn gaps into neutral reviewer questions.
Keep missing authority, freshness, substantiation, consent, or policy as
unknown. Ask the authorized owner which process governs the missing item; do not fill it from model inference.Route the decision to the owner who has it.
Sales leadership and operations own sales governance; product and legal/compliance owners control claims; HR and counsel control employment matters; privacy/records owners control data and retention; authorized humans control communications.
Stop at the review handoff.
Create a local
needs_reviewreceipt only. Do not create or alter a forecast, CRM record, target, claim, price, evaluation, plan, message, or public content.
Bounded agent roles
Three clerks. No decision-maker.
O*NET describes sales-manager responsibilities that include goals, training, sales statistics, people management, and external relationships.1 A label packet is not a forecast, coaching assessment, performance review, or authorization.
Safe versus approval-required
AI prepares labels. Authorized people decide and act.
| AI may prepare | Named human approval is required |
|---|---|
| Format fictional or organization-approved, minimum-necessary labels. | Access, retrieve, upload, query, analyze, or share CRM, customer, prospect, employee, compensation, contract, quote, pricing, call, email, or account data. |
| Preserve supplied source, owner, time/version, purpose, type, and unknown/conflict status. | Score or rank a lead, account, deal, rep, candidate, or employee; infer intent, performance, likelihood, or eligibility. |
| Draft neutral questions about gaps, staleness, conflict, authority, claim proof, or policy. | Make or alter a forecast, target, quota, territory, price, discount, claim, plan, evaluation, compensation, hiring, promotion, discipline, or personnel record. |
| Produce a local review-only receipt. | Send, publish, contact anyone, make a commitment, approve a tool/vendor, or take any external action. |
Failure modes
Five shortcuts that quietly become decisions.
| Mistake | Why it fails | Repair |
|---|---|---|
| “Which deal will close?” | It asks the model to predict and rank a live commercial outcome. | Keep only supplied labels and route any forecast methodology and judgment to authorized sales leadership. |
| “Score these reps for coaching.” | It becomes a people assessment and could affect employment. | Do not enter person data. Route any coaching/performance process to authorized leadership, HR, and counsel as required. |
| “Recommend a discount.” | Pricing and commitments can bind the business and customer. | Mark pricing authority unknown and route the decision through the current approval process. |
| “Turn this proof into an ad.” | Objective product claims need an appropriate evidence and approval path.3 | Preserve the supplied reference and ask the named claim/product/legal owner for review. |
| “Update the CRM and email the team.” | That is a data write and communication, not preparation. | Stop at the local receipt. An authorized human decides any system or communication action. |
Runnable local artifact
Forecast & Coaching Review Desk
Enter only fictional or organization-approved, minimum-necessary labels. This form stays in the browser and creates a review-only receipt. It does not send, save, retrieve, query, score, forecast, or connect to anything.
Primary sources
What this manual is built on.
- O*NET OnLine: Sales ManagersOccupation context; profile marked Bright Outlook Updated 2026. [1]
- NIST AI 600-1: Generative AI ProfileVoluntary, cross-sector generative-AI risk-management context; published 2024 and page updated April 2026. [2]
- FTC Advertising FAQ's: A Guide for Small BusinessPublic educational context for claim substantiation before ads run. [3]
- EEOC/DOJ: AI tools and disability discriminationOfficial context for employment-assessment risk; not individualized legal advice. [4]
Sources provide context, not permission to use private data, assess people or sales outcomes, make a decision, or take external action. Recheck current sources, organization policy, and applicable jurisdiction before non-fictional use.
Related guides
Where this guide fits.
Proposed links: the Market Research & Marketing Manual for evidence and claim review, Follow-Up Systems for human-owned next-action workflows, Sales Engineer Evidence Desk for technical discovery, Retail Sales Supervisor Brief Desk for frontline shift preparation, and the Occupation AI Workflow Guide Directory for the broader shelf. Use these related guides to keep each workflow's authority boundary clear.