gtm-analytics

GTM Analytics

Turn commercial data into a trustworthy decision. Use this skill to create a GTM measurement plan, investigate performance, define a funnel, or design an experiment—not to manufacture precision from incomplete data.

Begin with the decision and definitions

Identify the decision owner, decision to make, business model, time horizon, unit of analysis (person, account, opportunity, subscription), source systems, and known data limitations. Define every metric precisely: numerator, denominator, inclusion/exclusion rules, time window, grain, owner, and refresh cadence. Reconcile stage names and identity rules before comparing teams or time periods.

Build a measurement chain from business outcome to leading indicators. Use cohorts and segments when averages would hide important differences. Separate correlation, attribution, and causality. Attribution models are decision aids with bias, not proof of channel truth; present multiple views when the decision warrants it.

Design experiments and forecasts honestly

For experiments, specify the intervention, audience, hypothesis, primary metric, guardrails, duration, sample or practical feasibility, and decision rule before observing results. For forecasts, document inputs, conversion assumptions, lag, scenarios, uncertainty, and the point at which the forecast should be revised.

Deliverable

Return a metric dictionary, funnel or journey map, scorecard/dashboard specification, key findings, data-quality risks, and recommended actions. Provide formulas and sample queries only when requested or useful. Label missing data and assumptions explicitly; do not change reporting, tracking, CRM fields, or dashboards without approval.