Last December as Pacific Shaving Company's marketing department was prepping to send out their holiday sales campaign, President Stan Ades wondered how to make it perform better.
Frankly the design was pretty darn good (see sample below), with attractive graphics, easy-to-skim type, and a compelling free shipping offer. What else could Ades tweak as a test to raise response?CAMPAIGN
"I had spent the past two years testing send days/times for our emails in order to optimize my efforts as best I could," says Ades. "After analyzing the data, I found there was very little delta in my results -- the timing of my emails had very little effect on sales.
"Then in December, it hit me: don't just look at the gross revenue or number of sales transactions by campaign, but look at WHO is purchasing with each campaign."
When his team reviewed the past campaign data they discovered an overwhelming trend. "When we sent emails on weekends, an overwhelming number of responders had Web-based email accounts such as Yahoo. When we looked at the weekday sends and the customers who converted, the large majority of them were from business-based (corporate) email domains."
So, the team split the house list into two segments based on email domain. Names with the following, presumably at-home, addresses were mailed the campaign at 10 a.m. EST Saturday morning:
adelphia.net aol.com att.net bellsouth.net charter.net comcast.net cox.net dc.rr.com earthlink.net excite.com gmail.com hotmail.com mac.com mindspring.com msn.com netscape.net netzero.com nyc.rr.com optonline.net pacbell.net prodigy.net sbcglobal.net verizon.net yahoo.com
All other names at presumably at-work addresses were sent the exact same campaign at 2 p.m. EST the following Wednesday. Then Ades repeated the test using the same segments and timing in February for his big pre-Valentine's Day campaign last month.
The response lift far exceeded Ades' expectations. Although he didn't do an A/B test, comparing results to a control send, we feel the difference in year-over-year sales to what was basically the same offer sent to the list speaks for itself:
Holiday 2003 (entire list sent to Wed. 2 p.m. EST) 24.1% open rate 5.8% clickthrough 4.8% conversion to sale
Holiday 2004 -- At-home address weekend send 52.6% open rate 19.1% clickthrough 10% conversion to sale
Holiday 2004 -- At-work address weekday send 52.9% open rate 18.9% clickthrough 8% conversion to sale
Valentine's offer 2004 -- (entire list sent to Wed 2 p.m. EST) 32.8% open rate 8.5% clickthrough 5% conversion to sale
Valentine's offer 2005 -- At-home address weekend send 45% open rate 18.6% clickthrough 8% conversion to sale
Valentine's offer 2005 -- At-work address weekday send 47.1% open rate 18.2% clickthrough 7.8% conversion to sale
Although you may not experience an identical lift for a test like this, these numbers should convince you to at least test the concept.
Useful links related to this article
Creative samples of the two email campaigns: http://www.marketingsherpa.com/pacshav/study.html
Pacific Shaving Company http://www.pacificshaving.com