AI Lifecycle Marketing OS (Dual-CRM)
Built for a direct-to-consumer beauty and wellness brand
Overview
An intelligent marketing layer that sits on top of a direct-sales brand's commerce, email, SMS, and commission systems and coordinates them as one. It pairs two connected CRMs, a Customer CRM and an Ambassador CRM, each with a full behavioral data model and an automated trigger-driven messaging engine, plus a staff command center and a self-service ambassador portal.
The Problem
The brand ran promotions as manual calendar blasts on top of a brittle CSV handoff between its back-office CRM and its email tool, which caused missed and mistimed sends. Off-the-shelf tools handled basic customers but could not model the direct-sales side: acquisition attribution, the customer-to-ambassador pipeline, payout psychology, or service-aware suppression. Support and subscription data sat siloed from marketing, so promos sometimes reached people who had just had a bad experience.
My Approach
Architected a unified behavioral data model spanning roughly 80 customer fields and 60 ambassador fields, including computed signals like days-since-last-order, at-risk score, and percent-to-next-rank. Built a trigger-driven engine of about 70 flows across onboarding, retention, cart and browse recovery, win-back, subscription save, service recovery, and rank advancement, with acquisition source shaping the welcome and open support tickets suppressing promos. Added a payout-psychology sub-sequence that treats earning and receiving money as separate moments. Layered in AI-generated per-recipient copy, subject-line testing, and a rules-based queue that promotes high-value customers into the ambassador onboarding journey.