If we have 315k in a sending data extension and at time of scheduling use exclusion DEs (w/expected total send volume of ~295k) -- how can we implement send throttling to make sure the send takes about 2 hours?

My assumption (which turns out to be incorrect) was to setup an hourly threshold of 150k and the send would take a little over 2 hours. However, the send was complete after ~1 hour 15 minutes.

  • How did you measure the "complete after 1hour and 15 minutes"? Did you research the sendlog and _send dataview? Nov 25, 2020 at 6:47
  • 1
    can you check individual sending times? I would speculate that it did one send at the start of the throttling period, and a second one starting one hour after, both slowed down, taking around 15 minutes to complete. In the business rule notes from the days where support used to activate the feature, it says that Note: The send throttle will not always send at this rate consistently. , so I don't think you can expect an even spread. Nov 25, 2020 at 12:08
  • Turns out I had old information around how the MTA in marketing cloud processes an email using an hourly threshold. I was told it distributes the send in 12 five-minute increments throughout the hour.
    – Daniel
    Dec 7, 2020 at 20:12

1 Answer 1


This is covered in the help docs "Send Throttle"

The system begins processing the email send job at the start time you specify. The send is not spread evenly between the start and end times. The system continues sending during the specified time range until the send is complete or the end time you specified is reached, whichever comes first.

So send throttle starts at the top of the hour and sends till it hits that throttle. So if the system can send 150k in 15 mins it stops sending at that time and picks up at the next hour. It does not spread the send over the entire hour. If you look at the dataview _Sent you will see all the sends grouped near the top of the hour

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .