The best day to send email?

Latest posts | Feed | By Mark Brownlow

calendarAfter discussing the best time of day to send an email, it makes sense to look at the best day, too.

We all agree that the answer is "it depends": you need to pick out your best guesses based on what you know of your audience, emails and organization and then test to find the winning day.

Good.

But what if that process isn't practical or particularly insightful? And is the idea of a "best day of the week" even the right approach?

What do studies say?


Most experts are leery of those benchmark reports that aggregate numbers across thousands of senders to find which day of the week produced the best open or click rates.

After all, if you have a "what to do this weekend" newsletter, you're not going to send it on a Sunday morning just because that's when studies say the average list gets the best open rates. You don't have an "average" list.

Another argument against using aggregated benchmarks is that the reported "best" day keeps shifting around.

Here are some "best day of week to send" results (based on open rates and some measure of click rates)...

As reported by MailerMailer for 2004 - 2008:

Best day to send email table 1

As reported by eROI for intermittent periods from 2005 - 2007:

Best day to send email table 2

So, yes, the "best" day keeps changing. But note how often the weekend and start of the week pops up compared to, say, Thursdays. Surely no coincidence?

Does that mean we should always send sometime between Saturday and Monday?

No, because we still have the issue that your list and organization has its own unique characteristics that might make Thursdays the best day.

In fact that "what to do at the weekend" email might perform best on Thursdays, when people are beginning to plan their Saturday and Sunday.

But these aggregated stats do have value. Here's why...

First, if you really can't come up with any realistic "best guess" day, then benchmarks are a starting point...it's worth testing the weekend or early week for your emails.

Second, these stats bring home the point that you need to base your decision on the right metric. The reported best day for opens is rarely the same as the best day for clicks. And the best day for conversions may be different again. What are you trying to maximize?

Past performance and segmenting by daily responses


You can apply the same principles to finding the right day to send as we suggested for finding the right time. Namely:

1. Look back at your campaign reports and see if you can pick up on any patterns. Do emails sent on a particular day always get better results?

2. Define segments based not (just) on what people click, but also on which day they click.

Not everyone on your list has the same email habits. You may find some that check email on weekdays only. And others that use Sundays to catch up on email. Segmenting by "preferred day" lets you better time your sends to match the recipient's email habits.

Change the question: new thinking


All of the above and most articles on this topic fall into a common trap, however. Thanks to the legacy of a newspaper mindset, we often think of the "best day to send" issue in terms of days of the week.

But the question isn't "what's the best day of the week to send email?" It's "what's the best day to send email?"

There's an important difference.

Put it like this. You worked out Sundays are the best day of the week to send out your email promotions.

OK, Christmas Day is on a Monday.

Are you going to send your "gifts with free overnight shipping" promotion on Sunday, December 24th?

Of course not.

Whenever we send campaigns out related to specific events, weather, etc., we implicitly acknowledge that the "best day of the week" and the "best day" are not always going to be the same.

Once you think of the "best day" as independent of a weekly timetable, it opens up numerous possibilities.

One is to explore those trigger email campaigns that essentially let the recipient determine the best day to send. Because the email is sent relative to some event or recipient behavior. More on that here.

Another is to draw up a list of those factors that determine the best time for a recipient to take the action you want them to (click, view, download, buy, register etc.).

The choose your send date based on a review of those factors.

One such factor is when recipients check their email, which is where "best day of the week / best time of day" comes in. And that's still likely to be a critical factor, especially for informational publication-type emails.

But other factors are often undervalued, representing missed opportunities: especially for promotional emails.

Dela Quist writes, for example:

"A far greater number of people on your list are likely to buy because they have just been paid than the number of people who buy because you sent the email before 10am on a Wednesday."

Sending your email on a day of the month where people are likely to have cash in their pockets is a no brainer, as confirmed by Morgan Stewart.

Dela argues strongly in favor of looking at more critical timing factors that are independent of the day of the week, rather than searching for the artificial compromise that is a best day of week (or best time).

Payday is just one example of a factor that might drive the timing of a campaign. I'm sure you can come up with many more that are relevant to your list, organization, and emails.

Think time/day combinations


Another trap is to think of "best day" and "best time of day" as two separate issues. The best time to send depends on the day you send it. And vice versa.

This is illustrated in a detailed test involving B2C campaigns conducted by Switzerland's Newsmarketing agency.

They tested time/day combinations to come up with a matrix (see Page 12 of their report) that shows, for example, that Saturday evenings get great responses...but Monday afternoons are better than Saturday mornings. (NB: report is in German.)

This kind of detailed testing pays off if you can then build the associated time/day list segments and adjust the timing based on the multi-factor approach outlined above.

In summary, it's clear that simply finding some "best day of the week" or "best time of the day" to send your email is a worthy exercise, but only an initial step in optimizing your timing.

Any thoughts?

More on timing | Tags: ,

Permalink | June 09, 2009 | 4 comment(s)
Get posts like this: as an RSS feed | biweekly email | via Twitter

4 Comments:

ideal scenario, of course, is to do it by individual recipient. 'time of day' across the whole list becomes less valuable the larger your list is.

much harder to do than to say though!
By Anonymous dan barker, on 10 June, 2009  
 

"much harder to do than to say though!"

Indeed. Kind of hoping the technology folk will take care of that problem for us.
By Anonymous Mark Brownlow, on 10 June, 2009  
 

Hey Mark, nice comparison. I'd like to add a note... It's critical to test sending on different days with your own list. Senders who promote weekend events will find better responses on Fridays, when many people are finalizing weekend plans. Businesses who target other businesses might find their big day to be Mondays. While companies like ours (MailerMailer) report averages, we recommend that marketers take a closer look at their target demographics and behavior patterns to optimize what works best for them.
By Anonymous Raj Khera, on 10 June, 2009  
 

Thanks Raj, sound advice.
By Anonymous Mark Brownlow, on 11 June, 2009  
 

Comments closed for this post