PDM Workflows¶
AI-powered workflows for keyword analysis and paid media performance reporting.
/pdm-chatgpt-ads-readiness-pre¶
Assess whether a Seer client should run ChatGPT Ads using Scrunch data. Use for ChatGPT Ads readiness audits and AI-search bid/no-bid decisions; produce readiness scores, competitor benchmarks, a topic-level bid map, and an HTML dashboard. For recommendation-only analysis, use geo-aeo-recommendation-analysis.
/pdm-competitive-analysis¶
Paid media competitive intelligence: ad library extraction (Meta, LinkedIn, Google), auction insights analysis (new competitor detection, IS pressure, brand bidding), and messaging gap analysis. Called as a sub-skill by pdm-weekly-optimization Mode 2 signals 2.1 + 2.3 — receives competitor domain + auction context, returns structured brief with messaging gaps, creative patterns, and pitch-ready recommendations. Also runs standalone for competitor-launch deep-dives, pre-pitch briefs, share-of-voice analysis. Use when researching competitor strategies, analyzing ad copy patterns, finding messaging gaps, investigating new auction entrants, or building competitive intelligence for client conversations. Triggers: 'analyze competitor X', 'who is bidding on [client]', 'new competitor in auction', 'competitive brief', 'extract ads from [company]', 'competitive landscape'.
/pdm-keyword-analysis¶
Analyst-driven deep-dive deliverable for paid search keyword performance. Generates a full report with keyword performance tiers (CTR + CVR matrix), search term mapping, n-gram pattern analysis, and prioritized recommendations. Delegates negative-keyword identification, opportunity surfacing, and cross-campaign overlap detection to the pdm-sqa sub-skill (single source of truth for SQR data). Use when an analyst needs a complete keyword analysis deliverable for a client report or quarterly review, NOT for the automated weekly cadence (use pdm-weekly-optimization for that). Triggers on: 'run keyword analysis', 'deep-dive on keywords', 'keyword performance report', 'analyze paid search for [client]', 'quarterly keyword review'.
/pdm-methods¶
Paid media analysis, reporting, and optimization methods for Seer paid media work. Use when creating paid media reports, analyzing campaign performance, or applying paid media best practices.
/pdm-report¶
Generate paid media performance reports with executive summary, efficiency analysis, and actionable recommendations. Use when creating weekly/monthly paid media reports or analyzing campaign performance.
/pdm-sqa¶
Search query analysis for paid search accounts. Stateless sub-skill — takes explicit inputs, always returns the same structured output regardless of how it was called. Runs three analyses: negative keyword identification (zero-conversion queries segmented into auto-approve and AM-review tiers), new keyword opportunity surfacing (high-volume uncaptured queries), and cross-campaign query overlap detection. Covers Google Ads and Microsoft Ads. Called automatically by pdm-weekly-optimization for search accounts during Mode 1 health checks and Mode 2 opportunity signals. Also independently triggerable for: new campaign launch, post-match-type migration, client search term question, or standalone SQA request. REQUIRES org_id — all queries are scoped to a single client.
/pdm-testing¶
End-to-end paid media testing/experimentation framework with five modes: Mode 1 monthly AM-led roadmap planning (intake + cross-channel conflict checks + hypothesis generation + Wrike task creation), Mode 2 test registration on launch, Mode 3 daily automated monitoring with a deterministic decision engine (SRR + TCS + volume gates + KPI delta classification), Mode 4 wrap-up workflow (full test data + structured report + AM business-context verification + structured learnings capture), Mode 5 cross-client learnings search via anonymized shared index. Covers A/B, time-based, geo, and net-new experiment types across Google Ads, Meta, LinkedIn, YouTube, Microsoft Ads, Reddit, and programmatic. BigQuery-only via Seer MCP. 3-tier config resolution (Claude.ai Project Instructions, workspace config, or inline prompt). Use when planning new paid media tests, running daily test monitoring, wrapping up a completed test, capturing learnings, or searching past test outcomes across clients.
/pdm-weekly-optimization¶
Weekly paid media optimization workflow with two modes. Mode 1 runs weekly — executes the 7-category account health check, calls pdm-sqa for search accounts, writes findings into the Wrike weekly task description, and surfaces a 1–3 item focused agenda. Mode 2 runs daily — silently monitors 20 signals across 6 families (performance, competitive, efficiency, structural, opportunity, platform intelligence), fires only on threshold breach, opens a Claude conversation, and creates Wrike sub-tasks only when action items are confirmed. Configuration resolves via 3-tier pattern: Claude.ai Project Instructions (preferred), workspace config file (.seer/client.json), or inline prompt — supports Claude.ai, Claude Code, OpenCode, and Agent Skills Spec agents. Use when running weekly optimization, account health checks, or daily anomaly monitoring for paid media clients.
Skills¶
- pdm-methods — Paid media analysis, reporting, and optimization methods for Seer paid media work. Use when creating paid media reports, analyzing campaign performance, or applying paid media best practices.
Last generated: 2026-09-28 from 8 skills