P

AI Agent Role Playbooks

Performance Marketing Roles for agentic MCP

2 agentic playbooks for AI coding agents on Performance Marketing.

Available free v1.0.0 LLM
$ sidebutton install marketing
Download ZIP

Marketing Analyst & Media Buyer playbooks for Performance Marketing agents

2

Performance Marketing role playbooks run on the SideButton MCP server. Install sidebutton install marketing to enable these agent role playbooks for any coding agent.

Marketing Analyst

analyst

Measures campaign performance, builds attribution models, runs the Four Levers diagnostic, and converts raw ad data into bid/budget/creative/audience decisions. Turns "what happened?" into "do this next week." Every analysis task follows this pattern:

Media Buyer

media-buyer

Plans campaign structures, writes RSA/social ad copy, sets targeting and bidding, watches daily pacing, kills losers, and scales winners across search, social, display, and programmatic. Owns the buying side — what to run, where, for how much. Every campaign task follows this pattern: Load context — read media-context.md. Understand business goal, KPIs, budget, audience, and channel scope. If no context exists, ask for one before proceeding.Audit current state — before building anything new, assess what exists:Review active campaigns, their structure, and recent performanceIdentify budget waste (high spend, low conversion segments)Check tracking and attribution setupNote what's working (don't break winners)Plan campaign structure — design before building:Map campaigns to funnel stages (awareness → consideration → conversion → retention)Define campaign → ad group → ad hierarchyPlan audience segmentation and exclusionsSet budget allocation across campaignsChoose bidding strategy aligned to KPI (CPA target, ROAS target, max conversions)Build — create campaign components:Write ad copy following channel-specific best practicesDefine targeting (keywords, audiences, placements)Set bids and budgetsConfigure conversion tracking and UTM parametersEnsure landing page message matchLaunch and monitor — controlled rollout:Launch with conservative budgets, scale after validationMonitor learning phase (don't touch campaigns during platform learning)Check for disapprovals, tracking fires, budget pacingSet up automated alerts for anomaliesOptimize — data-driven improvements:Wait for statistical significance before making changesOptimize bids based on conversion data, not clicksPause underperformers, scale winnersTest new creatives, audiences, and copy systematicallyReallocate budget toward highest-performing segmentsReport — communicate results and next steps:Report against the KPIs defined in media contextInclude spend, conversions, CPA/ROAS, and trend directionRecommend specific next actions with expected impactFlag risks and opportunities