By Sean Donahue, Editor

For the past several weeks I’ve been traveling around the country teaching an email list growth session to groups of B2B marketers. During the most recent session one attendee asked me what was the best day of the week -- and time of day -- to send email.

My answer: It depends.

It’s not the most satisfying answer, of course. But the fact remains that there is no universal "best" time or day to send email. What works best for a B2C marketer sending ecommerce promos might generate lackluster results for a B2B marketer sending a newsletter or event invitation. And even a good-performing send time can begin showing weakness as your list evolves, subscribers’ preferences change, and external factors -- such as the economy -- alter behavior.

Testing -- and re-testing -- time of day and day of week is a must to determine your organization’s optimal send time. So we’ve compiled four test ideas from our case study archive that have helped marketers find meaningful improvements in their own send times.

Test Idea #1. The "sweet spot" for multiple time zones

If you can’t segment your list by location and stagger sends according to time zones, try finding a time that works for subscribers in broad range of locations.

For example, the team at MarketingExperiments (MarketingSherpa’s sister company) had seen improvements from sending their regular web clinic email invites at 6 a.m. rather than 9 a.m. EST. So, they wondered whether sending those emails even earlier might create a further lift.

The tested sending at 2 a.m. EST and achieved slight gains over their 9 a.m. send:
o 3% increase in open rate
o 1% increase in CTR

But the nature of those clicks proved interesting: They saw an increase in registrations from Europe and Asia -- where subscribers received the message during work hours. They also saw registrations from subscribers on the West Coast of the US, who were still on their computers at 11:00 p.m. PST.

On the other hand, Mike Atkinson, Director, Internet Marketing, House of Magnets, reported that his team has success sending later in the morning -- 10 a.m. PST. That way, he says, West Coast recipients have had time to clear out their inbox in the morning, while East Coast subscribers are just returning from lunch and are ready to settle back to work.

Test Idea #2. Workdays vs. weekends

B2C marketers often have to decide how to best reach their consumer audience: During business hours, or over the weekend?

One way to test subscriber preference is to segment your list according to corporate email address or web-based, at-home email clients such as Yahoo!, Hotmail or Gmail.

The team at skincare products retailer Pacific Shaving Company performed this type of segmentation, then tested different send times for two key holiday shopping campaigns (Christmas and Valentine’s Day):
o Corporate domain emails received messages on Wednesdays at 2 p.m. EST
o Web-based clients received messages on Saturdays at 10 a.m. EST

Comparing results from the previous years’ campaigns, they found that segmenting the list and staggering send times delivered major improvements over one send time for the entire list.

- Results from the segmented Christmas campaign:
o 121% increase in open rate
o 227% increase in CTR
o 87% increase in conversion rate

- Results from the segmented Valentine’s Day campaign:
o 40% increase in open rate
o 116% increase in CTR
o 58% increase in conversion rate

Test Idea #3. Day of signup

You might find clues about your subscribers’ email habits from their past activity with you.

For example, the subscription-based "brain games" site Lumosity wanted to improve site usage and paid conversions for members who had signed up for a free trial. They wondered whether sending reminder and conversion emails to people on standard days, such as Tuesdays or Thursdays, might be reaching people at times when they couldn’t go online to play.

So they looked back at their records to discover the day of the week on which subscribers had originally signed up for the trial. Then, they conducted an A/B test that compared sending emails to free trial members on the same day they signed up, vs. a control group who received emails on Tuesdays or Thursdays.

Response rate for the group who received email on the same day as signup increased significantly, with site usage also rising 81%.

Test Idea #4. Historic open and click activity

You can personalize email send times even further by analyzing your database and looking for trends in how individual subscribers respond to your messages.

The team at personal finance website Mint.com worked with their ESP to analyze data on past subscriber behavior to determine what time of day they were most likely to open and click. Then, they conducted a test to determine whether sending messages at this time would improve performance.

Half the list received emails at their personalized time based on past behavior. The other half received messages at 6 a.m. EST

Messages sent at personalized send times generated:
o 7% better open rates
o 13% better CTR

Useful links related to this article

Members Library -- Overnight Send Time for Email Lifts Open Rate, CTR, Registrations

Members Library -- Double Campaign Results by Sending Emails to At-Work & At-Home Email Addresses on Different Days

Members Library -- Best Day to Convert Trial Signups: Test Results

Members Library -- Behavior-based Email Send Times Lift Opens, CTRs and Referrals: Test and Results

MarketingExperiments

House of Magnets

Pacific Shaving Company

Lumosity

Mint.com