A/B test plan
Hypothesis. For a customer who cancelled their first appointment, removing friction beats adding persuasion. Variant B bets they were already sold and just need the fastest path back to a time. Variant A bets the cancellation reflects lingering doubt that needs reassurance: real groomers, included services, social proof.
What varies. One variable at the structural level: persuasion content vs. time slot shortcuts. Header, headline, footer, and send timing are identical, so a winner is attributable to the layout strategy, not the copy.
How it runs in Klaviyo. Native flow A/B on this message, 50/50 split. Flow conversion metric set to Appointment Booked with a 30 day window. Let it run to significance on conversions, not opens. At Barkbus flow volumes this likely means several weeks, so no peeking and no early calls.
Decision metric. Primary: rebooking conversion rate per recipient. Guardrails: unsubscribe and spam complaint rate, and click to booking completion rate to catch a variant that wins clicks but loses checkouts. Secondary: time to rebook.
What we learn either way. If B wins, the whole flow should get leaner and slot links should propagate to the SMS touch. If A wins, doubt is the real churn driver and email 2 should double down on trust content like groomer profiles and what to expect.