Revenue operations review

Revenue Operations Review

Produce one Revenue Operations Contract that turns customer evidence into a trustworthy, auditable flow from demand through qualification, solution proof, commitment, forecast, handoff, and outcome. Treat CRM as an operating ledger with scoped authority—not a spreadsheet-shaped source of wishful thinking.

Atomic boundary

Own revenue object semantics, lifecycle/pipeline states, evidence gates, attribution and routing, qualification/discovery, solution/technical handoff, forecast and capacity logic, incentives/commission control effects, automation, reconciliation, and change governance. Do not own campaign execution, pricing/package decisions, contract interpretation, billing/ledger truth, product implementation, or customer-success playbooks.

When not to use

• Use marketing-automation-blueprint, saas-subscription-pricing, or payment-platform-readiness when the primary artifact is demand generation, price/package choice, or payment/ledger correctness.
• Use enterprise-contract-operations-review or customer-success-operations-review when the primary artifact is executable contract obligations or recurring post-sale customer value rather than revenue pipeline truth.

Resource routing

• Read references/revenue-system-and-crm-truth.md for every task.
• Read references/qualification-forecast-handoff-and-incentives.md when discovery, qualification, solution proof, forecast, capacity, handoffs, quotas, commissions, or incentives matter.

Source verification

Retrieve current CRM/provider capabilities, source schemas, employment/commission requirements, privacy rules, contract policy, and accounting/forecast definitions at execution. Never hardcode vendor fields, legal rules, or claim pipeline outcomes not observed.

Operating rules

1. Label facts given, observed, customer-stated, seller-or-agent-asserted, assumed, hypothesis, or decision. Keep activity, evidence, stage, forecast category, commitment, signed contract, invoice, cash, and realized customer value distinct.
2. Define revenue motion, customer segments, buying/problem context, roles or autonomous agents, account hierarchy, products/offers, currencies/regions, sales/decision cycle, canonical systems, decision authorities, and downstream consumers before changing fields or dashboards.
3. Give each object and field one definition, owner, allowed writer, source, required state, quality rule, history, sensitivity, retention, and downstream dependency. CRM is canonical only for declared commercial workflow facts; product, contract, billing, finance, support, and customer-success truth remain with their owners.
4. Model lifecycle and opportunity states with entry evidence, exit evidence, permitted reversals, expiry/staleness, next action, owner/automation, and terminal reasons. Stage must reflect customer and evidence change, not internal activity volume or desired forecast.
5. Separate identity/account matching, source attribution, qualification, routing, engagement, opportunity, solution proof, commercial approval, contract, handoff, and outcome. A lead score or meeting cannot silently create a qualified opportunity.
6. Make discovery falsifiable: customer problem and impact, current alternative, affected users/workflow, stakeholders and authority, constraints, timing/trigger, desired outcome, evidence, risks, and disqualifiers. Preserve what is unknown and what the customer actually said.
7. Convert demos, POCs, integrations, security answers, roadmap requests, and bespoke commitments into typed evidence and handoffs with success criteria, caveats, owner, authority, due date, expiry, and customer-visible status. Never let notes become product or contract promises.
8. Forecast from explicit definitions, time-indexed snapshots, evidence, stage transition behavior, amount confidence, date confidence, scenarios, capacity constraints, and calibration. Separate pipeline creation, forecast, bookings/contract, billing, cash, and revenue recognition.
9. Design incentives and commissions from desired customer/business outcomes plus gaming analysis. Separate eligibility, credit allocation, splits, attainment, approval, adjustment, dispute, payout, clawback, and audit; retrieve current legal/employment authority and never let compensation fields redefine finance truth.
10. Automate reversible routine work now: capture, enrichment with provenance, dedupe, routing, evidence checks, next-action drafts, stale-state detection, forecast recomputation, handoff packets, reconciliation, and quality repair. Do not create low-value human or agent data-entry merely to feed reports.
11. Gate autonomous external actions by consent, approved claims, account/channel policy, authority, frequency, and risk. An agent may prepare evidence and execute bounded workflows; it may not invent customer intent, pricing, roadmap, legal/security claims, discounts, or binding commitments.
12. Select validation and recovery controls from the affected truth, downstream effect, reversibility, blast radius, and risk. A local presentation-only field change may need schema/consumer tests, narrow readback, and revert; material state, attribution, forecast, routing, capacity, incentive, or authority changes require the applicable shadow calculations, replay, migration/backfill, downstream impact, fairness, calibration, gaming, duplicate/conflict, provider-failure, audit, rollback, and live readback. Never optimize field completeness at the expense of truth or perform every control merely to complete a checklist.
13. Treat customer records, seller/agent assertions, forecasts, capacity, incentives, commissions, routing, and internal quality/abuse signals as protected operating evidence. A customer, partner, board, or public view is a separate purpose-bound, audience-named, minimum allowlisted projection with tenant/account authorization where applicable and negative leakage tests; it never inherits the internal CRM schema wholesale.

Workflow

1. Establish revenue truth and decisions

Map motion, segments, customer journey, actors/agents, canonical systems, object graph, current problems, decision users, metrics, authority, privacy, and ruin boundaries. Identify which facts are evidence, prediction, commitment, or outcome.

2. Design objects, states, and quality

Define account/contact/buying-group/opportunity/product/quote/handoff objects as applicable; field contracts; lifecycle and pipeline state machines; dedupe/merge; attribution; routing; permissions; history; quality; stale/unknown behavior; and downstream joins.

3. Design qualification and proof flow

Specify discovery evidence, disqualifiers, routing, technical validation, demo/POC success, security/procurement dependencies, commitment controls, commercial approvals, solution-to-contract and post-sale handoffs, and customer-safe communication.

4. Design forecast, capacity, and incentives

Define forecast taxonomy and horizons, amount/date/scenario uncertainty, calibration, pipeline coverage limits, delivery/implementation capacity constraints, incentive/commission rules, gaming and conflict analysis, disputes, corrections, and finance boundaries.

5. Automate, migrate, and assure

Create the automation authority, event/reconciliation contracts, and risk-selected migration/backfill, shadow/canary, alerting, change approval, dashboard, data-quality repair, audit, incident, rollback, and observed-outcome controls. Retire unused fields, stages, reports, and automations.

Hard gates

Reject or redesign an output that:

• has undefined objects, stages, amounts, dates, attribution, forecast categories, required evidence, allowed writers, or canonical owners;
• treats activity, seller confidence, lead score, meeting, proposal, verbal intent, signed contract, invoice, cash, and revenue as interchangeable;
• rewards stage inflation, overselling, bad-fit deals, hidden renewal pressure, low-quality pipeline, or commitments without authority;
• lets CRM notes create product, security, legal, pricing, contract, billing, or customer-success truth;
• publishes one forecast without scenarios, calibration, staleness, amount/date confidence, capacity, and change history;
• automates external outreach or commitments without source, consent, frequency, claim, privacy, and authority controls;
• changes fields, stages, routing, or incentives without classifying affected truth, downstream effect, reversibility, and risk, then applying the corresponding migration, downstream, replay/shadow, rollback, training/agent-contract, and reconciliation controls;
• claims forecast accuracy, attribution causality, pipeline health, capacity, or revenue outcome from invented or selectively cleaned data.

Output contract

Produce one Revenue Operations Contract containing:

1. artifact ID, scope, evidence labels, revenue motion, segments, actors/agents, canonical systems, authorities, decisions, definitions, ruin boundaries, audience/sensitivity, and authorized external projection contracts;
2. canonical object/relationship/field registry with writers, sources, quality, sensitivity, history, retention, and downstream consumers;
3. identity/account, attribution, lifecycle, pipeline, routing, dedupe/merge, stale/unknown, and terminal-reason state machines;
4. qualification/discovery evidence contract, disqualifiers, solution proof/POC, commitment controls, and typed handoff records;
5. forecast taxonomy, snapshots, scenarios, calibration, amount/date uncertainty, pipeline/capacity model, and decision rules;
6. incentive/commission eligibility, credit/split, attainment, approval, adjustment, dispute, payout boundary, gaming, and audit controls where applicable;
7. automation authority, privacy/consent, reconciliation, dashboards/metrics, data-quality, change-risk classification, applicable migration/backfill, change governance, incidents, and rollback;
8. validation/evidence plan with the applicable replay, shadow/canary, downstream tests, cohort and gaming review, live readback, explicit non-applicable dispositions, and specialist handoffs.

The contract is complete when every commercial state and prediction can be traced to evidence, no downstream system is forced to treat CRM opinion as fact, and automation can operate without inventing authority.