Drop Tara into Claude and get a Senior Predictive Marketing Analyst who turns model outputs into marketing actions — CLV frameworks, churn risk models, propensity scoring, platform-native ML setup for Klaviyo, Braze, and HubSpot, and holdout tests that prove the predictions actually work. No predictions without actions. Tara always pairs a model output with a corresponding marketing trigger. She translates churn risk percentages into revenue at risk before presenting to stakeholders — and always recommends platform-native ML before suggesting custom Python models. What you get → CLV framework — calculation methodology, segment architecture, and channel CLV comparison → Churn model design — leading indicators, scoring specification, threshold definitions, and interventions → Propensity scoring — next-purchase, upsell, and reactivation model specification → Platform prediction guide — configuring and activating Klaviyo, HubSpot, or Braze native ML → CLV by channel analysis — cohort LTV comparison framework that reframes channel budget allocation → Holdout test design — control group experiment to measure true incremental lift from predictions → Predictive dashboard spec — metrics, data sources, and visualisation design for marketing leadership 📄 tara-predictive-marketing.skill Under 5 min install Works with Claude, ChatGPT & any AI chat How to install Download the .skill package → open Claude → paste the file contents into your Project Instructions or system prompt → start working. That's it.
