Multi-Campus K-12 · Build
How a 75-campus school network fixed its enrollment funnel without buying more leads
The network had enough leads. They were going cold before anyone followed up. We built an AI agent that responds to every new inquiry within a minute with a message tailored to that family, proved it on a four-campus pilot, and rolled it out to every school in the network.
BuildThe problem
A K-12 network with about 75 campuses across the US was admitting families at well below its target rate. Leadership thought the problem was lead volume. The real problem was speed and quality of follow-up. The families who received sustained, personal follow-up were far more likely to enroll, but most leads got only a handful of touches, a meaningful share had never been contacted at all, and a large share had no source attribution because the booking tool silently dropped tracking parameters.
What we built
- New leads get a personalized text and email within a minute. AI reads the lead's source, the child's age, and any prior interactions, then writes a message that fits. Generic auto-responders sat at the bottom of the funnel. Personalized, source-aware ones are what the best-converting families were already getting.
- Legal guardrails are built in. State law bans marketing texts before 8am and after 9pm. The system enforces that automatically across all campuses. It also checks opt-out status and campus rules before sending anything.
- School directors walk into tours prepared. When a tour is booked, AI researches the family and writes a one-page brief before the meeting. Directors used to walk in cold.
- The leads that had never been contacted got personalized re-engagement messages sequenced over two weeks. Reaching that many families by hand was never going to happen. AI did it in days.
- A fix to the booking flow restored source attribution that was being lost on every booking. Leadership can finally see which marketing is working.
- A full attribution audit cleaned up duplicate lead sources in the CRM so the AI has accurate data to work from.