GEO Workflows¶
Generative Engine Optimization skills: build a client's LLM prompt set, analyze answer-engine recommendations, and produce the expanded GEO report. Plugin: plugins/divisions/geo.
/geo-aeo-recommendation-analysis¶
Run AEO Recommendation Analysis for any Seer client with Scrunch data. Measures whether the brand is actively recommended by LLMs (not just mentioned or cited) using a regex-proven 5-tier spectrum (T0 Invisible to T4 Recommendation). Outputs a single-file HTML report with tier distribution, platform and stage breakdowns, T4 quality, verbatim T4 exhibits, ghost citations with stranded brand equity, branded vs nonbranded comparison, negative and mixed sentiment exhibits, and strategic recommendations. Includes a single-window baseline mode for clients with under 60 days of data or recent collection cadence shifts. Use when someone asks to measure LLM recommendation rate, prove whether a brand is recommended in AI responses, run a Be Chosen analysis, analyze recommendation language, or evaluate whether the model is doing the selling vs just acknowledging the brand. Triggers on phrases like is X being recommended by ChatGPT/Claude/Perplexity or what does the model actually say about X.
/geo-expanded-report¶
Generate or update a Seer client's monthly expanded GEO, LLM visibility, or AI visibility report as interactive HTML. Use for client reporting, not a ChatGPT Ads bid/no-bid decision; use pdm-chatgpt-ads-readiness-pre for that.
/geo-prompt-builder¶
Build, expand, or audit GEO monitoring prompt sets for any Seer client on the Seer GEO Prompt Builder v3.6 methodology, then optionally push to Scrunch. Produces a Scrunch-ready pipe-delimited CSV sized per client from the Seer Prompt Tier Framework (100-prompt Tier 0 baseline for new and BD clients, scaling by decision tree), 80/20 unbranded/branded, persona-framed, nine intent types, bias-gated, with a deterministic QA validator covering structure, tagging, and grammar. Use when asked to build GEO prompts, create an LLM or AI visibility monitoring prompt set, size or refresh a client prompt set, or audit a supplied set for bias. Triggers: 'GEO prompts', 'prompt set', 'how many prompts', 'AI visibility prompts', 'Scrunch prompts', 'audit prompts for bias', 'prompt builder'. Upstream of geo-brand-canon-pre (Brand Canon) and geo-brand-accuracy-score-pre (response scoring); hands coverage-gap questions off to geo-prompt-coverage-gap-pre.
Last generated: 2026-09-28 from 3 skills