Segment your e-mail housefile

Sure, it’s time-consuming. It may double or even triple the time you spend putting together an e-mail campaign.

But according to industry experts, the potential sales and conversion benefits of segmenting your e-mail housefile—that is, sending different messages and offers to different segments of your customer file—are well worth the effort.

The experts also agree that e-mail housefile segmentation isn’t much different from segmenting your print buyers; it’s just a little more involved. “What you get with e-mail is more behavior-based information,” sums up Reggie Brady, president of consultancy Reggie Brady Marketing Solutions. “The trick is making sense of that information and organizing it into useful segments.”

Before You Begin

One of the bigger challenges of e-mail database segmentation presents itself even before segment categories are specified: determining whether all the customers in your database have truly opted-in to receive offers. This is crucial not only for maintaining good customer relations (and avoiding the appearance of spam), but for analyzing later the results of your campaign.

Opinions vary about whether customers who have provided their e-mail addresses via different channels should be considered true opt-ins. Pamela McAtee, director of e-marketing solutions at PreVision Marketing, contends that the possibility of receiving an e-mail offer is implicit when a customer gives a retailer his or her address via any channel.

Meanwhile, Ken Burke of e-consultancy Multimedia Live, believes that some less Internet-savvy customers may think the provision of an e-mail address when placing a Web order is just a tool for order verification, and not necessarily an invitation to receive offers. Try to avoid this confusion in the first place by making the e-mail registration page on your Web site as clear as possible. Or send a precautionary permission e-mail to ensure that your segmented campaign, once it’s deployed, isn’t misinterpreted by recipients as spam.

Once opt-in status has been determined, tighten your database further—de-duplicate it by e-mail address, says McAtee. She warns e-tailers against relying on customer identification numbers as the unique specifier in their e-mail database, as customers who use multiple channels may end up having multiple IDs. This can result in one customer receiving several e-mails with the same offer.

Sample Segments and Offers

Burke recommends that small catalogers with in-house e-mail operations who have never segmented before start by running a simple A/B e-mail split test to just two segments and see how each responds. “It’s not all that sexy,” he admits, “but it’s what people should be doing.”

Or, says Reggie Brady, try following the 80/20 dictum of purchasing: Segment the most profitable 20 percent of your e-mail customers and craft offers only for them.

When narrowing it further, you may find the following e-mail segments among your largest and most common:

- Product-related. This covers categories such as frequency of purchase, date of last purchase and type of product bought.

- Offer-related. What types of offers have these consumers responded to in the past?

- Geographic location. Discover the customers who live near your retail stores, then send offers that will drive them there. According to a case study released by DoubleClick, Crate and Barrel saw store sales increase by 50 percent from the previous week by doing this.

- Channel-related, or whether these customers came from your Web site, retail store or print catalog.

Other sub-segments and offers also can be tailored around your merchandise. Burke cites the example of a jewelry e-tailer that discovered that segmenting its offers by gender yielded success. Or, he says, an electronics retailer may want to customize separate offers for computer, audio and video buyers.

No matter what form it takes, it’s important that you discern the No. 1 criteria within your customer base that makes sense to isolate, and then e-mail that segment accordingly. Besides free shipping or a free gift, very few offers will succeed across the board.

Outsource When Possible

Because of the time and technological might often needed to clean and segment a list, many catalogers choose to outsource this function along with the deployment of their e-mail campaigns.

This carries several advantages. For example, outsourcers often have built-in safeguards that let you specify your criteria for designating an opt-in customer, saving you the trouble of designating these customers yourself. Among some of the major third-party e-mail deployment providers are Bigfoot, Digital Impact and ClickAction.

If you choose to keep your e-mail deployment and segmenting in-house, be sure your IT and marketing staffs are sufficiently trained and ready for the task. “If you don’t have the technology to do this,” warns Burke, “you’re really out of luck.”

Companies that provide in-house e-mail applications with segmenting functions include DoubleClick and E.piphany.

Analyzing Results

Whether they’re spent in-house or out-, your segmenting efforts ultimately may be thwarted if you don’t properly analyze the end results of your split campaigns. To ensure that you do this accurately, embed each clickable area of your e-mail campaign with a traceable source code. When these areas are clicked, the data is logged and stored for later study.

Burke often uses an additional square-inch analysis when evaluating an HTML e-mail campaign. He looks at the productivity of each area of the e-mail, and adds or eliminates areas according to his findings.

Also, Brady encourages e-tailers to analyze the results of numerous segmented campaigns taking place over several months before forming conclusions about the quality or even the rate of each segment’s response.

Caveats

Avoid the temptation to over-segment your e-mail database; this makes it difficult or sometimes impossible to track the resulting campaign data. If you’re working with a housefile of 100,000 names, says Brady, 100 segments might be overkill, but five would be manageable and easy to track.

Also be wary of certain domain names. Brady says customers whose e-mail addresses belong to free providers such as Yahoo! and Hotmail often use this address as a catch-all for junk e-mail. In other words, they may never check the account, so they may not be worth segmenting out.

Remember these caveats, but don’t let them stop you from divvying up your e-mail database to your best advantage. As Brady says, “The biggest mistake is not segmenting at all.”