Skip to main content
reopt Handbook
reopt Handbook
New Brand Marketing Strategy

Brand Strategy

Brand Strategy

Agent-Based Marketing

Agent ArchitectureMarket IntelligencePersona Engineering

Digital Marketing Channels

Search MarketingContent MarketingSocial MediaPerformance Marketing

Customer Relationships and Offline

CRM and EmailOffline Marketing

Measurement and Operations

Marketing AnalyticsExperimentationCalendar and OKRs

References

Verification
Handbook›New Brand Marketing Strategy›Agent Architecture
한국어English

Agent Architecture

Design AI-assisted marketing workflows with clear roles, inputs, and review gates.

Key takeaways

  • The architecture defines what AI agents may draft, what humans must approve, and what evidence every output cites.
  • Five agent roles (research, persona, content, performance, analytics) each have a required input and a human review gate.
  • Agents never invent customer evidence, metrics, or legal claims; brand voice and forbidden claims live in a shared context pack.
  • High-risk channels like paid ads and pricing pages require explicit approval before launch.
  • Agent value is measured by decision speed and asset quality, not raw content volume.

AI agents should increase marketing throughput without turning the brand into generic output. The architecture must define what agents can draft, what humans approve, and what evidence every output must use.

Agent Roles

AgentMain jobRequired inputHuman gate
Research agentSummarize market, customer, and competitor signalsSource list and research questionSource quality review
Persona agentConvert signals into audience hypothesesSegment, job, pains, triggersStrategy review
Content agentDraft outlines, posts, emails, and landing copyBrief, voice rules, proof pointsEditorial approval
Performance agentSuggest creative tests and budget movesCampaign data and constraintsSpend approval
Analytics agentExplain KPI movement and anomaliesDashboard export and contextDecision review

Workflow Pattern

Operating Rules

  • Agents never invent customer evidence, case studies, metrics, or legal claims.
  • Every draft includes the source, target persona, funnel stage, and intended action.
  • Brand voice and forbidden claims live in a shared context pack.
  • High-risk channels such as paid ads, pricing pages, and legal-sensitive claims require explicit approval.
  • Agent output is measured by decision speed and asset quality, not raw content volume.

Useful Artifacts

ArtifactOwnerUpdate cadence
Brand context packMarketing leadMonthly or after positioning change
Prompt libraryChannel ownerWeekly during campaign cycles
Approved claim bankMarketing and salesAfter each new proof asset
Review rubricEditorial ownerQuarterly

Related docs

Content Marketing

Design content pillars, formats, and reuse loops across the buyer journey.

References

Reference categories for operating a brand marketing system.

Context Pack

Agentic MVP · Package the information an AI agent needs to build the MVP slice safely.

AI-Era GTM

A practical handbook for AI-assisted go-to-market strategy and revenue operations in B2B SaaS.

Multi-Agent Architecture

Vercel Enterprise AI Platform · Design supervisor, delegation, pipeline, and review patterns for enterprise agents.

Brand Strategy

Define the strategic foundation that makes marketing decisions consistent.

Market Intelligence

Build a repeatable research loop for customer, competitor, category, and channel signals.

On this page

Agent RolesWorkflow PatternOperating RulesUseful Artifacts