Obsessed with B2B marketing? You should be a PRO member! Join now at 25% off (or 50% off for teams).

Understanding customer behavior is key to creating marketing campaigns that generate high response and revenue.

One of the best ways to understand customer behavior is to study customer migration patterns—to learn when and why a customer ends up in a segment different from the one he or she had been in.

The starting point for those studies is your customer-segmentation model. After you decide which approach to use to measure migration, the process is a virtuous circle of analyze, segment, campaign, and analyze again.

The next task, requiring strong analysis skills, is to tie the observed migrations to company activities, such as a marketing stream, and to customer purchase behavior.

The final task is to apply the results of that analysis to your marketing campaigns to generate revenue and boost customer retention.

Why Study Migration Patterns?

Customers are not statues, cast in stone and plunked down in a park as a perch for birds. Some buy a lot, and some defect to the competition.

Though you likely have some very steady buyers, most customers are in motion. That is not a new notion. We wrote about it four years ago in the Target Marketing magazine article "The Five Laws of Velocity Marketing." Today marketers refer to that behavior as customer migration.

We study customer-migration patterns because they help us understand customer behavior and ultimately increase sales.

We want to know why some customers defect and others remain loyal. We want to know how to allocate the marketing budget to get more customers moving up in our rankings than there are customers moving down.

We want to sell more goods and services. We want to maximize our revenue and marketing return on investment (ROI). What we learn from migration patterns can help with all those tasks.

The Role of Customer Segmentation

Migration implies movement from here to there. In marketing applications, the "here" and the "there," where a customer started and where the customer is now, respectively, are defined by the customer groups, or "segments," to which the customer is assigned.

The metrics used to define segments and the values of those metrics, which differentiate one segment from another, come from an analysis of the purchase transactions of the customer population.

The process of making the segment assignments is called "customer segmentation." Choose whichever metrics serve your marketing goals at the time. A customer's migration to a new segment usually suggests a marketing opportunity.

Customers could be segmented by their body weight (which might be important for a specialty-clothing retailer), by how they respond to a survey, by demographics such as ZIP code, or by their buying behaviors.

The most useful and predictive way to segment customers is by buying behaviors. Even choosing to segment by buying behaviors leaves a lot of wiggle room. Perhaps the most popular behavioral metric is Recency—how many days since a customer's last purchase.

Some companies use revenue, dividing customers into groups according to how much they have spent over some reasonable period. Another frequently used methodology is RFM (a combination of Recency, Frequency of purchase, and Monetary value, or revenue).

We use a sophisticated measure called Loyalty Rank and explain in another whitepaper, RFM vs. Loyalty Builders' Modeling why that metric is more accurate and more predictive than others.

Obviously, the set of metrics that is used for the segmentation can affect the segment into which a customer will fall and how the migration patterns will look. Since our ultimate goal is to increase revenue, the objective is to choose a segmentation model that supports an effective, differentiated contact strategy.

By marketing to each segment according to its characteristics, we are able to craft the best message for each customer group and maximize total revenue.

Two Ways to Measure Customer Migration

There are two ways to approach the measurement process. In the first, think of the segments as a set of buckets. Count the population in each bucket at the start of the period being studied and then count the population in each bucket at the end of the period.

Differences between the counts reflect migration by segment. Figure 1 is a real-life example of a migration report using the first approach.

This company has seven segments:

  • Loyalists
  • Nurture
  • Underperforms
  • Fading
  • Winback
  • Onetime and two-time buyers
  • Inactive

The best customers are the Loyalists. The lowest-ranking customers are in the Inactive segment. Typically, one-time and two-time buyers are pulled out and treated separately, but they are a very important segment that contains potentially excellent customers as well assome who are just passing through.

The segments in the example are defined by customer scores on two metrics: Loyalty Rank (sometimes called LRank) and Risk Score (sometimes called Risk Probability); check out definitions in this glossary.

Analysis and segmentation occur monthly, and the population in each segment is counted monthly, with the most recent month at the bottom. The cells show how the segment population has changed compared with the previous month.

For example, in August 2009 the number of customers in the best segment, Loyalists, dropped by 2,179 compared with the number of Loyalists in July. Apparently, many customers moved down one segment to Nurture, for that segment gained 3,136 customers. Also, the top group and the third group (Underperforms) apparently have been losing members for the past four periods, and the second group has been gaining members over that same period.

The report is easy to calculate and at first glance seems to show some meaningful information. Did the gain in the Nurture group come more from improving Underperforms or sinking Loyalists? Did the last month's drop in one-time and two-time buyers come from a decrease in acquisition activities, or did those customers just go inactive? It is hard to answer those questions using the bucket approach.

Enter your email address to continue reading

Segment Migration: Where Your Customers Were, Where They Went, and Why

Don't worry...it's free!

Already a member? Sign in now.

Sign in with your preferred account, below.

Did you like this article?
Know someone who would enjoy it too? Share with your friends, free of charge, no sign up required! Simply share this link, and they will get instant access…
  • Copy Link

  • Email

  • Twitter

  • Facebook

  • Pinterest

  • Linkedin


ABOUT THE AUTHOR

Dr. Mark Klein is CEO and founder of Loyalty Builders LLC (loyaltybuilders.com) and three other companies. He blogs frequently on mathematical marketing and recently published his first novel. You can reach him at 603-610-8800.