Most marketers are already well aware that to maximize the response to customer marketing campaigns, it is critical to tailor the message and offer to each individual customer, or small groups of very similar customers. When a customer receives a personalized and highly-relevant offer, there is much more chance that the message will resonate with them. When marketers succeed in customizing their messages and incentives so that they are the most relevant and interesting for each individual customer, they can maximize revenues from their existing customers.
In order to accomplish this, the first step is effectively segmenting customers into groups which share similar behavioral histories. It is necessary to identify and isolate segments of customers with similar product preferences, purchase frequency, spending levels and so forth. In many cases, how and when the customer found the business in first place is an important predictor of how they will behave in the future. In some cases, demographic details (e.g., gender, geographical location, age) may also play a part. Each business is different; the goal must be to segment customers into small groups that reflect the way they will likely respond to campaigns. Sophisticated marketers and retention experts have discovered that identifying smaller, more homogenous groups enables far more targeted – and profitable – campaigns.
The Finer the Segmentation, the Better
Most marketers begin segmentation efforts by focusing on the large, high-level segments which are easy to identify. For example, segmenting customers based on their current lifecycle stage (e.g., New, Active and Churn) will yield a small handful of segments which can be used to begin tailoring marketing. Another example used by many marketers is segmenting their most valuable customers into different levels (e.g., Diamond, Gold, Silver, Bronze). Sending relevant campaigns to each of these high-level segments will yield higher revenues than sending the same campaign to the entire customer list. However, if the marketer can further divide these high-level segments into more granular sub-groups, they can expect to achieve even higher revenues.
For example, let’s take the group of Active customers and segment it into three sub-groups, based on the average per-order spend levels of each group: Low (average spend of $50 or less), Medium ($50-200) and High (more than $200).
(This example – based only on average purchase amount – is a deliberately simplistic one in order to demonstrate the concept. In actuality, you will want to combine multiple segmentation layers in order to identify distinct “customer personas” using cluster analysis and target them together. An example of such a persona might be “high-spending women’s apparel purchasers with a high risk of churn.” The most useful customer personas are those which identify a small segment of similar customers for whom the marketer can develop highly-relevant offers/incentives.)
The next step is to create tailored campaigns designed to maximize response according to each group’s individual behaviors. In the following example, the marketer’s goal is to increase the average purchase amount of customers in each group. For example, sending the Active-Low customers (whose average spend is $50 or less) an incentive of, “Spend $70 on your next purchase, get a 20% discount” will likely generate a relatively high response rate, because the deposit level is slightly above the group’s average. On the other hand, sending the same offer to the Active-High group (whose average purchase amount is more than $200) might actually reduce their spend amounts and cause a loss of revenue. Conversely, offering the Active-Low group, “Get a 30% discount on orders over $500” will not likely lead to a high response rate because this purchase level is far above their typical purchase level. But sending this offer to the Active-High group may encourage a good number of them to spend more than their typical average deposit amounts. The following diagram shows the results of sending customized campaigns to these three customer segments.
When marketers do this successfully, they typically achieve a much greater total uplift from the separate campaigns to the various sub-groups than they would achieve by sending the same campaign to all customers.
(It is important to note that “campaign uplift” can only be accurately measured by using test and control groups for marketing campaigns. Here is a good blog post discussing this: How to Treat Every Customer Campaign as a Marketing Experiment.)
The goal is clear: to continually define more and more granular customer segments to which to send the most relevant and enticing incentives. Maximizing the revenues generated by customer retention campaigns comes from optimizing the matching between many small sub-segments and the most effective messages and incentives for each.
Pini Yakuel – CEO and Founder of Optimove
Pini Yakuel has over a decade of experience in analytics-driven customer marketing, business consulting and sales. Pini is founder and CEO of Optimove, a profitable and rapidly-growing start-up based in Tel Aviv. His passion for understanding what drives customer behavior led him to spearhead the development of Optimove, an advanced retention automation platform powered by predictive micro-segmentation technology. Optimove is used by marketers and retention experts in dozens of Internet businesses in a variety of verticals.