All posts

Notes Sep 24, 2026 5 min read

What 1.1 million morning emails taught us about send time

We sent at 5:00 AM for four months and it nearly sank the idea. What six weeks of testing said about when an email replaces a habit, and what to do on the nine mornings a year Apple is late.

Casey Morgan Co-founder, CEO

A phone lying face down beside a mug of black coffee on a sunlit kitchen table

Ascent has sent 1,104,318 morning emails since the first one went out in March 2022. Every one of them had a choice baked into it that we made early and nearly got wrong: when it arrives.

The obvious answer was “as early as possible.” Apple’s daily reports for the US storefront are usually ready before 5:00 AM Pacific, and our importer finishes within a few minutes of that. For the first four months we sent the moment the numbers were in. Open rates were fine. Replies were not. People wrote to say the email was buried under everything else that arrived overnight, or that they read it at 5:40 AM in bed and then could not get back to sleep because downloads were down 12%.

Seven beat five, six and eight

In the summer of 2022 we split new accounts across four send times for six weeks: 5:00, 6:00, 7:00 and 8:00 AM, each in the reader’s own time zone. We measured three things: whether the email was opened that day, how long after sending it was opened, and whether the reader opened App Store Connect anyway before noon. We ask that last one in a one-question survey, once.

  • 5:00 AM was opened by 71% of readers the same day, a median of 2 hours 14 minutes after it arrived.
  • 6:00 AM reached 78%, at a median of 1 hour 20 minutes.
  • 7:00 AM reached 86%, at a median of 19 minutes.
  • 8:00 AM reached 83%, at 11 minutes, but 4 in 10 of those readers had already checked App Store Connect.

Seven o’clock is late enough that most people are awake and early enough that they have not gone looking yet. That second half matters more than the open rate. The email only earns its place if it replaces the checking. An email that confirms what you already looked up is just more email.

Your time zone, not ours

We are in Portland, so the first version sent at 7:00 AM Pacific to everyone. A developer in Raleigh got it at 10:00 AM, by which point the whole premise had expired. Sending in the reader’s zone sounds like a small change. It meant running the send once for each US zone from Eastern through Hawaii, and deciding what to do when Apple’s reports are late.

They are late about nine mornings a year. Our rule is simple: if the numbers are not in by 6:45 AM in your zone, the email waits up to 90 minutes and says so in the first line. If they are still missing after that, you get the email anyway with yesterday’s reviews and a plain note that downloads and proceeds will follow. We would rather send an honest half than a silent nothing, because a missing email makes people open the dashboard, and then we have lost the morning.

What we compare against

The other choice inside every email is the comparison. “612 downloads” means little alone. For the first year we compared with the day before, which made every Monday look like a triumph and every Saturday look like a crisis. App downloads have a weekly shape, and for most categories it is strong.

Now each figure is compared with the same weekday last week. A Tuesday is measured against the previous Tuesday. It is a dull choice and it removed about half the alarmed replies we used to get, because the swings it reports are real ones.

What we would tell you to copy

If you send your own users a daily or weekly summary, three things carried over from our numbers to the handful of customers who tried them:

  1. Send in the reader’s time zone, at an hour when they are up but have not started.
  2. Compare like with like. Same weekday, and the same week of the month if your product has a billing cycle.
  3. When the data is late, say so in the first line. Silence is read as an outage.

You can change your own send time under Settings, then Morning email. Most people leave it at 7:00 AM. The 6% who move it nearly all move it earlier, and we suspect they have small children.