Catering Customer Churn Prediction
Flags catering accounts at risk of not rebooking.
Model version: churn-prediction-v0.5
Methodology contact: Head of Catering Data
Start free trial — no credit card
ROI
Pilot in progress
N=14 trucks, started 2026-01-15, readout expected 2026-07-31. We will not publish a number until we have one.
What it does
For operators with repeat catering customers, scores each account on rebooking probability over the next 90 days. Surfaces a needs-outreach list each Monday morning.
Who it is for
- Catering Heavy
- Multi Truck
Integrations
What this feature doesn't do
- The model sees rebooking patterns; it does not see why a customer left.
- Outreach to a flagged account is a conversation, not a discount carpet-bomb.
- Some customers do not rebook because they actively disliked the service.
Pricing
Included with Operator Pro.
See all plansFAQ
- How long does Catering Customer Churn Prediction take to learn my truck?
- Roughly 30 days. The model needs to see a full cycle of weekdays, weekends, and weather patterns before its recommendations stabilize.
- What does Catering Customer Churn Prediction cost?
- Pricing is transparent on the pricing page. Most features are included with Operator Pro. There is no demo gate on headline price.
- Is the model version pinned?
- Yes. Catering Customer Churn Prediction runs on model version churn-prediction-v0.5. We disclose this on every page and in JSON-LD softwareVersion.
- What integrations does it require?
- Required integrations: mailchimp, klaviyo.
- What does this feature not do?
- Honest caveats: The model sees rebooking patterns; it does not see why a customer left. Outreach to a flagged account is a conversation, not a discount carpet-bomb.