Does Orita's engagement model take into account how frequently customers in a specific category typically purchase?

Yes. Every Orita customer's model is fundamentally different. Orita's ML is dynamic and adapts to each brand's specific customer behavior, analyzing the full history of events in Klaviyo rather than applying a fixed time window.

For brands where customers make large, infrequent purchases (like furniture), the model weights engagement signals very differently than for brands with frequent repeat purchases (like supplements). The model recalibrates continuously as more data flows in.