Predictive send time is a machine learning technique that forecasts the specific moment an individual subscriber is most likely to open an email, based on that subscriber's own historical behavior patterns, and schedules delivery accordingly.
A working answer to what is predictive send time worth keeping in mind: it’s a per-subscriber prediction, not a list-wide schedule decision, which is what makes it fundamentally different from picking a single best send hour for everyone.
The Mechanism Behind a Concept Already Covered
Send time optimization, covered under behavior based scheduling elsewhere in this glossary, is the broader outcome, delivering email at the hour a subscriber is most likely to engage. Predictive send time vs send time optimization comes down to level of description: predictive send time is specifically the technical mechanism that makes the broader outcome possible.
- Send time optimization describes the practice and the result, better-timed delivery.
- Predictive send time describes the underlying model, a system trained on open and click timestamps across a subscriber’s history, predicting the single best delivery window for that specific person.
The prediction happens per individual subscriber, not per list or per segment, which is what allows one campaign to actually deliver at different times for different people while still going out as a single scheduled send from the sender’s perspective.
How the Prediction Actually Works
A predictive send time model typically trains on a subscriber’s own open history, when they’ve opened emails in the past, how consistently, across what days and hours, and uses that pattern to forecast future behavior. New subscribers without enough history yet usually default to a broader, list-wide pattern until the system accumulates enough individual data to predict confidently for that specific person.
Predictive Send Time Examples
- A subscriber who consistently opens email around 7am before work receives campaigns scheduled to land right before that window, rather than at a generic 10am send time applied to the whole list.
- A subscriber active mostly on weekend evenings receives the same campaign hours later than the 7am subscriber, timed to their own actual pattern instead.
Related terms:
Predictive send time email delivery depends entirely on having enough historical open data to train against, which is why the benefit compounds the longer a store has been sending consistently. Adflipr’s automation workflows use behavior based scheduling as a foundation, with the option to layer predictive timing on top as a store’s send history builds up.
