marketing-analytics
Sets up, audits, and reports on marketing measurement — tracking plans, event schemas, attribution models, and the dashboards built on them. Use this to instrument a site or product, audit tracking nobody trusts, choose or interpret an attribution model, build reporting that answers a specific question, or reconcile numbers that disagree between tools.
git clone --depth 1 https://github.com/cbrock84/headcount /tmp/marketing-analytics && cp -r /tmp/marketing-analytics/plugins/demand-generation/skills/marketing-analytics ~/.claude/skills/marketing-analyticsSKILL.md
# Marketing analytics ## The tracking plan comes first Dashboards built on bad instrumentation are confident and wrong, which is worse than having none. Define, in writing, before implementing: every event, when it fires, its properties and their types, and the question each one exists to answer. An event with no question behind it is noise that will be maintained forever. Naming convention decided once and enforced: `object_action`, lowercase, past tense. Inconsistent naming is unfixable later without breaking historical data. ## Auditing existing tracking Numbers nobody trusts usually come from one of: - **Double-firing** on route changes in single-page apps. - **Events that stopped** when someone changed a selector or a component. - **Definition drift** — two tools counting "signup" at different moments. - **Bot and internal traffic** never filtered out. - **Consent and blockers** removing a meaningful and non-random share of data. Verify by doing the action yourself and watching the event arrive with the properties you expect. Not by reading the dashboard. ## Attribution Every model is wrong in a known direction. Pick deliberately and state the bias: - **Last-touch** — over-credits closing channels: brand search, retargeting. Under-credits everything that created demand. - **First-touch** — the mirror image; over-credits discovery. - **Multi-touch** — better, and dependent on complete tracking you probably do not have. - **Incrementality testing** — the only method that answers "would this have happened anyway." The most expensive and the most trustworthy. Use one model consistently for decisions, and check it periodically against a holdout. Switching models to make a channel look better is how organizations mislead themselves. ## Reporting Every report answers one question for one audience. Reports built to display everything get read by nobody. Show the metric, its comparison period, and the decision it informs. A number with no comparison is not information. Where a number moved, the report should say why or say that the cause is unknown — "unknown" is a legitimate and useful finding. ## Tooling Product and web analytics: Google Analytics 4, Amplitude, Mixpanel, PostHog, Plausible, and similar. Attribution: HubSpot or Salesforce campaign reporting, Dreamdata, Rockerbox, and similar. All of them model rather than observe — treat the output as directional and say so when you present it. Warehouse-native reporting beats a vendor dashboard the moment you need to join spend to revenue on your own definitions. ## Never - Add tracking before the plan names the events and their properties. Retrofitting a schema onto live data is a migration, not an edit. - Report an attribution number without saying which model produced it. The same period looks like different businesses under first and last touch. - Change an event definition without versioning it. Every historical comparison silently becomes wrong. - Build a dashboard nobody named a decision for.
Concentrates marketing and sales effort on a named set of accounts rather than on volume — qualifying whether the model fits your economics at all, building the account list and the buying group inside each, tiering effort against account value, coordinating so the account experiences one campaign rather than several, and measuring account progression instead of leads. Use this to decide whether to run an account-based program, build one, or work out why an existing one produces activity and no pipeline.
Designs orchestrator-and-subagent hierarchies for a repository — splitting agents by exclusive write surface, pairing every producer with an independent auditor, and enforcing the split with a script that runs in CI. Use this whenever the user wants to set up, expand, audit, or fix a multi-agent or subagent structure for a codebase; asks how to divide work between agents; wants agent charters, roles, or a surface map written; or is hitting agents that collide on the same files, review their own work, or drift from their remit. Also use when sizing a roster or deciding whether a new agent is justified.
Governs models and AI systems in production — intended use, evaluation, monitoring, human oversight, documentation, and the decision to deploy or retire. Use this before deploying a model or AI feature, when defining evaluation criteria, when a model's behavior has drifted, when assessing AI risk or regulatory exposure, or when deciding whether an AI system is fit for a consequential decision.
Produces executive-level research — market sizing, competitor mapping, trend analysis, and strategic intelligence — grounded in cited sources with the confidence in each claim made explicit. Use this to analyze a market or industry, map competitors, evaluate a market-entry or build-versus-buy decision, produce a research brief, or assemble evidence for a decision. Also use when comparing options that need a structured, evidence-based verdict rather than an opinion.
Optimizes for AI assistants and AI-generated answers — being retrievable, being cited, and being represented accurately when a model answers on your behalf. Use this when traffic is shifting from links to AI answers, when a brand is misrepresented or absent in AI responses, when planning content for retrieval rather than ranking, or when deciding how AI search changes an existing SEO program.
Improves visibility and conversion in the App Store and Google Play — metadata, keywords, screenshots, ratings, and the listing experience that turns an impression into an install. Use this to audit or optimize an app listing, plan a launch listing, diagnose poor install conversion, or improve store search visibility.
Protects and restores data — backup coverage and scope, retention, immutability against ransomware, and proving restores actually work. Use this to design a backup regime, verify restores, plan retention, protect backups from ransomware, or recover from data loss.
Runs the planning cycle — annual budget, rolling forecast, consolidation of business unit inputs, and the variance analysis that explains actuals against plan. Use this to build or challenge a budget, run a forecast cycle, consolidate departmental submissions, explain why results diverged from plan, or improve forecast accuracy.