Operate with human accountability
Who owns this on Monday, and what existing work stops?
Start within current functions. Map the work, decision, authority and capacity before considering reporting lines. Shared definitions and controls can coexist with local customer knowledge, craft and execution. A central standard does not require a central approval queue for every campaign. sourcesOPS §§16–26AI Augmented GTM Operating Model and Transformation Plan — Operating model; derived design, not independent organisational evidence.RT §§41–48AI Augmented GTM Red Team Verdict — Red Team; claim audit, investment challenge and simplification.Evidence for this chapter · 4 claim records
Three accountability cards
| Accountability | Purpose and decisions | Human work | AI-assisted or automated work | Skills and success measures |
|---|---|---|---|---|
| Commercial owner | Own customer outcome, intervention and scale/stop decision within actual authority | Trade-offs, budget/delivery negotiation, customer judgement | Evidence briefs, scenario explanation and routine follow-up | Commercial judgement; net contribution, customer guardrails and executed decisions |
| Finance and method partners | Own cost definitions and credible inference respectively; challenge the readout | Scope, counterfactual, uncertainty, benefit recognition | Queries, tested calculations and optional readout drafting | Finance/measurement competence; reconciled economics and valid decisions |
| Operations and technical owner | Own reliable delivery, permissions, support and retirement | Release/incident judgement, maintenance and recovery | Native workflows, monitoring, exact checks and receipts | Workflow/security capability; reliability, cost, review burden and recovery |
These are accountabilities, not a prescription for three new jobs. A small team may combine responsibilities, with independent challenge where exposure requires it. Suppliers can contribute skills but cannot assume the client’s business accountability. Creative, Product, Sales and Service retain their craft and domain decisions.
T4 records who recommends, decides, executes and approves each consequential action. Where several approvals are needed, specify them and still name the final accountable decision owner. Give reviewers enough time and evidence to reject a proposal. Include absence cover. If review volume exceeds capacity, reduce scope or redesign the workflow; do not rely on ceremonial approval.
A decision-led cadence
| Cadence | Work and output | Avoid |
|---|---|---|
| Daily where needed | Exceptions, broken delivery, current safeguards and containment | Reopening strategy on ordinary noise |
| Weekly | One-page outcome, diagnosis, alternatives, action, owner and readout | Slide production without a decision |
| Monthly or at a mature readout | Finance reconciliation, benefit realisation and feasible tranche choices | Moving budgets during valid experiments without a method decision |
| Quarterly or on material change | Strategic assumptions, evidence reuse, ownership cost and retirement | Adding new forums without closing old ones |
Cadences are configurable. A small firm can combine reviews. Every new recurring review replaces existing work or has an explicit capacity cost. Train through the selected workflow: practitioners verify evidence, managers practise real decisions, operators rehearse failures and Finance checks actual benefits.
Incentives matter. Reward customer quality, useful execution and honest evidence; do not pay analysts for positive tests or teams for agent counts. An agency saving requires compatible commercial terms. A sponsor demonstrates support by resolving a real resource conflict, not merely signing a charter.
Persistent pods, shared platforms or a Centre of Enablement require a durable backlog, funded authority and a net reduction in coordination burden. If no existing work can stop and no additional capacity is available, shrink the programme.
Related
Evidence for this chapter4claims reviewed, with confidence ratings, limits and sources.
Claim classifications
Accepted means a principle or mechanism survived critical review, not that it guarantees a return. Accepted with conditions means prerequisites apply. Qualified means the original claim has been narrowed. Advanced needs substantial readiness. Experimental needs a capped comparison. Unresolved means the evidence cannot settle the claim. Rejected means the recommendation is excluded.
High, Medium and Low describe confidence in the stated target, not the chance of a positive ROI. For a Qualified or Rejected historical claim, confidence applies to the decision to narrow or reject it and to the supported limit—not to the original headline. Read the headline and classification together. An accounting identity or design principle is not empirical evidence of a commercial outcome. Follow the linked chapters, Plays and tools to see where each claim applies.