Build a content performance hypothesis engine from past campaign data

Write the business questions this monthly analysis must answer before exporting any data. Specify the primary outcomes, such as qualified conversions, influenced contacts, opportunity creation, pipeline contribution, retention support, or organic demand capture, and separate them from diagnostic metrics such as sessions or CTR. Set the analysis period, comparison period, attribution lookback, timezone, currency, and minimum data thresholds. Define which content types, domains, languages, and lifecycle stages are included or excluded. Assign a data owner for GA4, Search Console, HubSpot, Airtable, and the final decision meeting. Publish these rules as a versioned analysis specification so month-to-month changes are intentional rather than accidental. Review and approve the specification with analytics and campaign owners before any export begins.

Create an Airtable table with one row per canonical content asset rather than one row per URL variant. Include asset_id, canonical_url, normalized_path, title, content_type, topic_cluster, primary_keyword, persona, funnel_stage, intended_goal, primary_CTA, campaign, author, publish_date, last_refresh_date, status, locale, owner, and notes. Add fields for GA4 landing-page key, Search Console page key, HubSpot page or campaign key, and redirect target so later joins do not rely on title matching. Use controlled select values for content type, persona, stage, goal, and status to avoid spelling drift. Identify duplicates, translated variants, parameterized URLs, and retired pages during setup. Have content operations approve the inventory before performance data is joined.