April 29, 2008
Can segmenting email to boost relevancy help you weather an economic downturn? Take a look at this Case Study.
An eretailer has been using gender- and store-specific email promotions to lift revenue and grow their customer mailing list. Includes five easy-to-follow tactics and results from a day-of-the-week test.
With consumers tightening their spending habits, many email marketers will be happy to see any kind of sales increase this year. James Connell, Senior Director for Ecommerce, Digital Marketing & New Media, Roots, a Canadian-based clothing multichannel retailer, realized late last year that his company might be affected, too. That’s why he and his team decided it was a perfect time to boost the relevancy of their emails.
“We needed to do a much better job of driving people to something that’s relevant,” Connell says. “I had been signing up for our competitors’ emails, and I had been getting mostly messages that didn’t apply to me. That was educational.”
Connell saw more relevancy as a competitive advantage their brand could pounce on. And with response rates dropping and consumer spending falling, he wanted to test some new ideas to optimize their email program.
Connell and his team were already following best practices on privacy and deliverability, but they were essentially sending the same offer to their whole email list over and over again. They knew the next step should be list segmentation … and time was of the essence because of the gloomy economic forecast.
First, Connell began by identifying some key areas where they could improve email relevancy.
o Targeting women and men with gender-specific offers
o Segmenting email lists even further based on their 128 store locations
o Identifying which days of the week were best for email by testing *every* day -- not just the industry-proven midweek days
Here are the five major tactics Connell and his team followed:
-> Tactic #1. Segment customer list by store location and gender
To deliver more targeted messages, Connell organized their list names into separate files according to store locations. They took the ZIP Code attached to each name from past online purchases and then assigned it to the closest store.
Then, he and his team divided the customer list even further into separate files for men and women. They manually went through their customer list to make the new files. For customers without distinctly male or female first names and purchase patterns, they created another file they labeled “undetermined.”
-> Tactic #2. Identify ‘undetermineds’
They didn’t stop there in segmenting their customer list. Every ‘undetermined’ name was emailed a survey to get seven data points:
o Last purchase date at Roots.com
o Last purchase date in a Roots factory outlet store
o Favorite department
o If they were interested in Roots’ yoga department
The survey of “undetermineds” allowed Connell and his team to further segment their customer files by gender and their 124 Canadian and 4 US stores.
-> Tactic #3. Promote signups at each store
The survey process also involved creating email signup forms for each store. New customers were offered the chance to win various prizes if they signed up. Store managers turned the names into Connell and his team by shop and city location.
-> Tactic #4. Create segmented offers
In the end, customers received both store- and gender-specific offers that involved printing out an email coupon -- usually a discount on an item -- to take to the nearby outlet.
- The copy pointed out that the offer was good not only online but also at their local Roots store.
- The address and a directions link were included to encourage redemption at the stores.
If a customer walked into the store without the email coupon but still mentioned the promotion, the stores gave them the same discount. Each store manager was responsible for reporting the number of redemptions back to Connell by phone or email after the promotion expired. There was no digitally computable coupon code.
Customers who were not within driving distance of the stores did not receive these offers. They were put into a separate online-only customer file. They received gender-specific offers that were designed to purely drive online sales.
-> Tactic #5. Test emailing *every* day of the week
Connell didn’t want to leave any money on the table by taking for granted the industry’s almost de facto ‘best days’ [Monday through Thursday] for emails. So, they tested every day of the week, including Saturday and Sunday.
Connell has no doubt that the segmentation tactics have increased the relevancy of the clothier’s email campaigns. At the same time, email-based sales have increased by 14% compared to the same period the year before.
“We’re happy to be getting a lift in these difficult economical times and expect the percent to continue to grow as we do more segmentation. We are seeing higher opens, clickthroughs and better ROI. And because of the increased relevance, our emails get forwarded more often and we are increasing our subscriber base.”
Connell and his team grew their subscriber base by 22% during the first six weeks of the campaign with the in-store email component. “About half the stores have participated in the email promotions so far, and that number is increasing.”
They also saw “a steady 40%” completion rate of the surveys sent to clarify a customer's gender and local store and they’ve been getting a 15% redemption rate for the email-exclusive offers as reported by the store managers.
Giving discounts to people who mentioned the email offer -- without proof that they received an email -- also appears to have been the right move. “We encourage email signups in the store, so it makes sense to accept the promotions in that setting to help create more of a multichannel identity.”
Connell couldn’t offer specific numbers on the best days for emails. But he says Tuesday and Wednesday proved to be their best send days. Still, team members were intrigued by the numbers they received on the Sunday tests.
“We are seeing surprisingly high clickthroughs on Sunday emails. There’s a lag time in sends for that day where people open up and click on Monday and even Tuesday. With the clickthroughs you see midweek, they happen very quickly or they don’t happen at all.”
Useful links related to this article
Creative samples from Roots’ email segmentation tests:
Sitebrand - provided marketing services: