Most businesses define their VIPs according to arbitrary delineations. In this blog post, we present a dynamic method for determining which customers are VIPs that bypasses the shortcomings of traditional VIP definitions.
All businesses know the importance of their big spenders, and marketers must have a well-structured strategy in place to keep their top customers satisfied and engaged. The importance of VIPs is overshadowed only by the difficulty of defining them. Delineating the VIP group is a constant struggle. On the one hand, underestimation will result in the loss of important customers and customer value. On the other hand, overestimation will result in squandered marketing dollars.
Most businesses define their VIPs according to a fixed spend or activity amount, or by fixed percentiles (e.g., the top 5% of customers according to a given metric). Sometimes, the cut-off point is determined according to the staff and resources available for VIP management.
These methods are lacking in two ways. First, they are arbitrary. They don’t account for the relationship between the VIP group and the rest of the customer base. And even more troubling – they are rigid in the sense that they don’t take into consideration the dynamic nature of businesses. For example, if you’re an e-tailer with a fixed spend-threshold VIP group and you’ve just introduced a line of pricey high fashion, your defined VIP group may quickly lose its applicability: you may suddenly find that a significant chunk of your customers suddenly become VIPs according to the predefined threshold.
An Easy, Dynamic Method to Define your VIPs
Let’s take a look at a better, more dynamic method for finding the optimal cutoff point for VIPs, one that overcomes the shortcomings of traditional VIP definitions. It’s not as accurate and dynamic as building a customer model based on predictive micro-segmentation (which requires dedicated customer marketing software), but it’s more data-driven than many prevalent alternatives.
This method uses a LN transformation and trendline to identify data outliers. The percentiles which violate the trend at its extreme are those which do not adhere to what we would expect of them under “normal” circumstances. This is the dynamic group of outliers which make up your VIPs. This method enables you to define your VIPs while taking into account the dynamics of your complete customer base and the variability of your business.
Here’s how to go about it:
1. After choosing a customer metric, divide your active customer base into 100 percentiles of average values, from lowest to highest (using only positive values larger than 1). The 100th percentile will have the highest value.
2. Calculate the values’ Logarithm (“LN”).
3. Use a scatterplot to graph the percentiles (X) and LN values (Y).
4. Add a trendline and format it as polynomial of Order 2.
5. The dots at the extreme right of the scatterplot violate the trend and are thus your “outliers,” your VIPs.
Shauli Rozen is a tech and data expert, consultant and business leader, specializing in creating business value through data-driven marketing and monetization strategies. His previous positions include Director of Strategy at Amdocs and Management Consultant with the Boston Consulting Group, where he advised senior management of Fortune 500 companies. Shauli holds a bachelor’s degree in computer science and an MBA from the Wharton Business School in the University of Pennsylvania.
At one time (not so long ago), most retention marketers would send the same marketing campaign to their entire customer list. Nowadays, most marketers understand well the value of segmenting their customers into groups and sending different offers to each group. As we’ve discussed in a previous post, sophisticated marketers and retention experts have discovered … Continued
At Optimove, we’re always thinking about ways to better understand customers and their behavior. A big part of that is constantly being on the lookout for interesting customer segments and personas that have the potential to improve customer retention marketing. One customer lifecycle stage that we think about a lot in the world of online … Continued
Many types of online businesses have customers who tend to be active only during particular times of the year. Just a few examples are online retail stores (many customers use particular stores to shop only for particular types of seasonal products, mostly ignoring the store the rest of the year), sports betting sites (many customers … Continued