Key Takeaways
- Inside Nova Intelligence, Send Time Decisioning sends each message at an hour chosen for each person.
Inside Nova Intelligence, Send Time Decisioning sends each message at an hour chosen for each person.
- Our setup guide lists the signals it reads: opens for email and push, clicks for SMS.
Our setup guide lists the signals it reads: opens for email and push, clicks for SMS.
- People with thin history get your scheduled time, and Quiet Hours take precedence over any chosen hour.
People with thin history get your scheduled time, and Quiet Hours take precedence over any chosen hour.
- Measure lift against a control group, or run a 50/50 A/B testing experiment.
Measure lift against a control group, or run a 50/50 A/B testing experiment.
A single send time treats your whole audience as one person. Send Time Decisioning, part of Nova Intelligence, chooses the hour each person is most likely to engage and sends then. Nova Intelligence is our AI layer, built natively into your data and journeys.
How Send Time Decisioning Works
Send Time Decisioning, part of Nova Intelligence, turns each person’s past behavior into a send hour. It works in three steps:
- Input: It reads past engagement in the same project, using opens for email and push notifications and clicks for SMS.
Input: It reads past engagement in the same project, using opens for email and push notifications and clicks for SMS.
- Decision: It chooses the hour each person is most likely to engage within the send window you set.
Decision: It chooses the hour each person is most likely to engage within the send window you set.
- Action: It sends the message at the top of that hour.
Action: It sends the message at the top of that hour.
A send window is the span of hours you give a send to finish. We update the models weekly and weight the newest engagement data most heavily, as our feature documentation explains. When someone’s habits change, the hour we choose for them changes too.
For example, imagine a weekly newsletter that goes to everyone at 10 a.m. With a window of 8 a.m.–2 p.m., a commuter who reads on the train gets it at 8:00 a.m. A lunchtime reader gets it at noon, and the 10 a.m. crowd keeps its usual slot.
What Send Time Decisioning Needs and What It Will Not Do
Send Time Decisioning works from past engagement, and it respects the delivery rules you already set. Five documented limits shape its results:
- Thin history: People with less than three months of engagement data get the campaign’s configured send time.
Thin history: People with less than three months of engagement data get the campaign’s configured send time.
- Quiet Hours: Quiet Hours, your blocked sending hours, take precedence. We hold the message until they end, or skip it if you set them to skip.
Quiet Hours: Quiet Hours, your blocked sending hours, take precedence. We hold the message until they end, or skip it if you set them to skip.
- SMS tracked links: SMS learns from clicks, so your SMS templates need shortened, tracked links. Without them, it has no SMS signal to learn from.
SMS tracked links: SMS learns from clicks, so your SMS templates need shortened, tracked links. Without them, it has no SMS signal to learn from.
- Fixed API send times: When an API request includes a scheduled send time (sendAt), Send Time Decisioning doesn’t apply to that send. See triggered campaign docs.
Fixed API send times: When an API request includes a scheduled send time (sendAt), Send Time Decisioning doesn’t apply to that send. See triggered campaign docs.
- No guaranteed lift: It may improve open and click rates, but our AI overview notes it can’t guarantee higher engagement.
No guaranteed lift: It may improve open and click rates, but our AI overview notes it can’t guarantee higher engagement.
Review these five before you judge results, since each one changes which sends the model can influence.
For example, a welcome series reaches mostly new subscribers, so expect many of its sends to use the scheduled time.
How You Measure the Lift
Send Time Decisioning reports its own results, so you can check whether individual send hours beat a fixed schedule. In Insights, you compare a treatment group against a control group: people who receive the message without an optimized send time.
What you see
What it tells you
Open lift
The gain in opens for optimized sends versus unoptimized ones
Click lift
The gain in clicks, the steadier signal for email
Top and least preferred send windows
The hours your audience engages with most and least
30-, 60-, and 90-day views
Whether lift holds steady across the past 30, 60, or 90 days
50/50 experiment
A fixed, even split between control and variation that you can’t adjust, set up like A/B testing
You find all five views under Insights > Intelligent Insights, as our Insights reporting docs show. Read them next to your standard open and click rates.
On May 27, 2026, we published customer medians compared with randomized send times. That baseline differs from a schedule your team already tuned, so compare against your own results too. Your lift may differ from these medians:
- The median US click lift was 73%.
The median US click lift was 73%.
- The median US open lift was 82%.
The median US open lift was 82%.
- The median EU click lift was 51%.
The median EU click lift was 51%.
- The median open lift in our EU results was 114%.
The median open lift in our EU results was 114%.
According to Litmus email client data, Apple Mail Privacy Protection now affects roughly 55–60% of email opens. Those opens don’t reliably show reading, so treat click lift as the steadier signal.
For example, 27% higher open rates versus control marked GreenFi’s first campaign with Send Time Decisioning.
Where Send Time Decisioning Sits in Nova Intelligence
Send Time Decisioning answers one question: when. Within Nova Intelligence, the Nova Decisioning grouping covers three capabilities, each making one delivery choice per person.
Capability
Decision it makes
What it uses
Send Time Decisioning
When to send
Past opens and clicks in the same project
Channel Decisioning
Where to send
Each person’s engagement patterns
Frequency Decisioning
How often to send
The engagement limits you set
Send Time Decisioning works in blast and triggered campaigns and in journey send tiles. See where it works in our help center. Together, the three capabilities extend real-time personalization beyond what you say, covering when, where, and how often you say it.
Start with Send Time Decisioning when timing is your open question. Add the other two as you see where channel choice or message volume holds results back.
Frequently Asked Questions
1. What is Iterable Send Time Decisioning?
In Iterable, Send Time Decisioning replaces one batch send time with an hour chosen for each person. You stop debating a single “best time to send” for your whole list. Your program also keeps adjusting as habits shift from week to week.
2. Is Send Time Optimization (STO) the same as Send Time Decisioning?
Yes. Iterable renamed Send Time Optimization (STO) to Send Time Decisioning on Oct. 5, 2026, and describes the change as a naming update tied to Nova Intelligence. Older help articles and some in-app labels may still use the former name, so search for both when you look for guidance.
3. Which channels does Iterable Send Time Decisioning cover?
Iterable Send Time Decisioning covers email, push, and SMS. SMS support began on Oct. 6, 2026. You enable it separately for each campaign and journey tile, so turning it on for email doesn’t switch it on for text messages.
4. What happens when Iterable has no engagement history for someone?
In Iterable, a person without enough history receives the message at the campaign’s configured send time. That fallback becomes the default experience for new
5. How do you know Iterable Send Time Decisioning works?
In Iterable, you run a send time experiment and compare the two groups in Insights. Start with a recurring send, such as a weekly newsletter, so you have repeated sends to compare. Judge the outcome on clicks first, because Apple Mail Privacy Protection makes many opens unreliable. Give the result time to settle before you decide whether to keep it on.
Put Send Time Decisioning to Work
One fixed send time can’t fit every person on your list. Send Time Decisioning picks an hour per person, and the lift report shows you whether it pays off.
- Follow our setup guide to turn it on for a campaign or journey tile.
Follow our setup guide to turn it on for a campaign or journey tile.
- Take the platform tour to see it alongside the rest of Nova Intelligence.
Take the platform tour to see it alongside the rest of Nova Intelligence.
