August 13, 2007

E-Mail Campaign Management And Data

The vendor community has many suggestions for improving e-mail marketing campaign performance.

Data plays an important role in determining which version of an e-mail campaign a customer should receive. There are several different attributes that should be considered.
  • Lifetime Purchase History. This is often the most important attribute in determining which version of an e-mail campaign a customer should receive. Marketers will categorize purchases by merchandise division, and use the merchandise divisions to determine which version of a e-mail campaign a customer receives.
  • Most Recent Purchase. In many instances, what a customer purchased five years ago no longer has relevance to what a customer will purchase tomorrow. Consider giving more weight to recent orders, giving less weight to older orders. Some marketers give purchases in the past three months a weight of 1.00, purchases 4-12 months ago a weight of 0.50, purchases 13-24 months ago a weight of 0.25, and all older purchases a weight of 0.125. If you have a statistician, you'll delight her with the task of assigning these weights!
  • Clickstream Data. What a customer looked at, especially in the past month, can be very relevant to determining which version of an e-mail a customer receives. Some marketers use shopping cart data, others categorize items viewed by merchandise division.
  • User Preferences. At Nordstrom, we asked customers what type of e-mail the customer wants (mens, womens, petites, plus-size, sale, etc.). We balanced preferences with purchases, weighting each differently when determining which version of an e-mail a customer received.
  • Demographics And Lifestyle Attributes. Traditional marketers like these factors. It can be easy to conceptualize a campaign for a 35-44 year old woman who owns a Lexus SUV. These attributes are often less powerful than the purchase history of a customer.
If you ever want to unite your marketing teams over a single idea, e-mail marketing is one place to do that. To execute an exceptional e-mail marketing program, with many targeted versions of an e-mail sent to different customers during one campaign, you'll want to leverage these folks:
  • Catalog Marketers, who have excellent experience determining "who" receives versions of a catalog.
  • Database Marketers, who have integrated data from all channels to provide a complete view of the customer.
  • Web Analytics Gurus, who can help summarize clickstream data for the Database Marketing team.
  • E-Mail Marketers, who oversee the entire process, are accountable for the end result, and manage vendor relationships, and understand the promotions/subject-line/template stuff that makes e-mail campaigns tick.
  • Traditional Marketers can help with the creative presentation most likely to be effective with customers.
  • Online Producers need to put the merchandise on the website, and make sure that landing pages work properly.
Sometimes all of the data nuances can become overwhelming. Performance can still be improved by simplifying the number of data dimensions, or by reducing the number of versions. Your e-mail delivery vendor may have the resources to help you with version assignment, if your budget provides you these opportunities.

Don't expect miracles from targeted e-mail versions. In reality, you have limited data for eighty percent of your e-mail file, so you won't do a great job of targeting to these folks. Among the top twenty percent of your e-mail file, these folks are so productive that many different versions of an e-mail campaign can work. If you can get a fifteen to thirty-five percent improvement in total campaign performance by targeting, you're well on your way to success.

4 comments:

  1. Anonymous5:27 PM

    Thanks for sharing your ideas on email marketing. Email marketing to an existing customer base can achieve great results from targetting based on customer needs. These go beyond demographics and clickstream tracking. When we really begin to understand why our customer is interested in our product, and we can track that to an individual customer, we can offer that person an extremely relevant email, at the right time.

    Only by truly understanding customer needs can we move into exceptional response rates.

    From your experience, Kevin, what do you find are "best practice" rates for email campaigns?

    ReplyDelete
  2. The folks at retailemail.blogspot.com do a good job of tracking open rates, click-through rates and conversion rates by industry.

    In my industry, you're lucky if six or seven percent of the customers who receive the e-mail actually choose to visit the website to browse (a twenty percent open rate times a thirty percent click-through rate).

    So you could triple that rate, and you still only get one in five people to visit the website.

    Worse, once the customer gets to the website, only three or four percent will buy something.

    The beauty of e-mail is that it is essentially free.

    Thre drawback is that it is essentially free, which lowers the desire of folks to work hard at targeting. In catalog, when you pay a buck to mail a catalog, you really work hard on the targeting side of the equation.

    ReplyDelete
  3. Anonymous11:08 AM

    For a lot of online & dot-com companies, they have behavioral data beyond purchase history that they can use to segment- such as login behavior, member activity, or special activities based on their kind of business. Using these to segment can push the clicks/opens up even more.

    Thanks for the writeup- interesting read!

    ReplyDelete
  4. Anna, good comments. Those are attributes that I omitted, thanks for adding them!!

    ReplyDelete

Note: Only a member of this blog may post a comment.

Two Articles For You To Think About

First, translate everything in this article about AI and Media to "AI and E-Commerce". Then you'll be interested in the topic ...