The best time to send your email?

Latest posts | Feed | By Mark Brownlow on July 11, 2007

clockeROI just released their latest email marketing stats on opens and clicks by day of the week and time of day.

The specifics are in the report, which you can download here. The lessons that jumped out for me:

1. If you've tested or made assumptions about the best day to send, redo the test or revisit the assumptions on a regular basis to account for changes in user behavior.

For example, weekends performed less well in Q2, 2007 than in the recent past. eROI suggest it's a seasonal change due to the better weather.

2. It's perhaps time to get a little bit more sophisticated about send times (notwithstanding limits imposed by bandwidth/volume constraints.)

We tend to default to sending all the emails out in one go. But what about the alternatives?

If you have a proven "best time for people to get your email" then stagger the send out by timezone (if you have appropriate information on your subscribers) so that people really do get it at that time.

Tests can be misleading here. Let's think of a list with subscribers in London, Los Angeles and Tokyo. If you send out all your emails at the same time, you can never hit the plum spots in all three cities.

Testing would likely show that the best time to send to the whole list in one go is a compromise: early afternoon in LA, late night in London, early morning in Tokyo.

Much better to send out your email in three batches: say when it's early afternoon in all three cities.

Even better, segment your list by timezone and then test each timezone for the best time to send. Working hours and email reading habits likely vary regionally.

Then how about personalizing the send time? Bill Nussey wrote about the success of this tactic a while back here.

What about sending email at the same time of day that the subscriber originally opted in?

Or sending email 10 minutes before the time of day they last opened an email from you? (And on the same day of the week.)

Not my original ideas: I know at least one ESP (Nussey's Silverpop) has the functionality to do the above. And I'm sure there are others.

Or how about linking sending times to some other subscriber behavior? A visit to an offline store? A call to customer service? A visit to the website (part of the premise behind trigger emails)?

Any ideas or comments on how else we might get a bit more clever with send times?

More on timing and frequency | Tags:

[This post brought to you by Campaigner Email Marketing]
Permalink | Add to del.icio.us | 2 comment(s) - add yours!
Get posts like this: as an RSS feed | as a biweekly newsletter
Sign-up for the Email Marketing Reports NEWSLETTER
Twice a month, free, packed with email marketing advice and all the posts from this blog.
Email:      First Name:     
    More info and sample | Your privacy is guaranteed

2 Comments:

Hey, Mark. I wouldn't recommend timing your emails to when a person subscribed. That's reading too much into a one-time event. I spoke with Shera Shrago, the head of email at Home Depot, a couple of months ago about this very issue and she suggested tracking open times a looking for patterns. If you can micro-manage it to that level, that may pay some dividends. I would also urge folks not to send email overnight, which is when most spam is sent. You don't want your emails getting caught up in the morning's spam.
By Blogger Chad White, on 13 July, 2007  
 

Thanks Chad, always good to get some frontline advice.
By Blogger Mark Brownlow - Email Marketing Reports, on 13 July, 2007  
 

Post a Comment