TL;DR: put your data to work and get actionable insights to drive profitable customer behavior by going beyond “the basics”.

Most Loyalty and CRM platforms provide a host of standardized reports and metrics: active members, customer lifetime value, new enrollments, response rates, etc.  Certainly, these standard reports and metrics are helpful, but to truly drive customer behavior it’s important to dig a little deeper.

So many companies rely on the same table-stakes metrics in an attempt to run and manage their loyalty program. However, in our experience, these baseline reports fall far short of what is necessary to run a truly “smart” CRM  and loyalty efforts.

In the following pages, we’ll walk you through each of these 8 analysis we employ to dig a little deeper.  We illustrated these with concrete examples using data from a fictitious restaurant chain.

ANALYSIS #1: Transaction Distribution

A transaction distribution helps you understand how often your customer base is purchasing from you.

  • Is your customer base comprised of ‘rabid fans’ who are all repeat customers?
  • Does your customer base have relatively few super high-value customers and a large number of “one-and-outers?”

A transaction distribution can identify whether you have an 80/20 business or a 60/40 business. Knowing that percentage difference, should dictate how you structure and communicate your loyalty program. For example, understanding the composition mix of the customer base in turn drives a host of marketing decisions.

If you operate a loyalty program, comparing the transaction distribution of members versus non-members is an indicator of whether your loyalty program is driving additional sales / visits, beyond levels seen in customers who aren’t part of the loyalty program.

Diving one level deeper, breaking out a transaction distribution by loyalty redeemers and non-redeemers can show how much lift and generative  value that redemption offered your business. When loyalty members redeem their rewards, does that redemption drive additional purchases that they wouldn’t typically have made?

TIP: From here, an easy add-on metric to calculate is average trips per customer, which is a fundamental metric for CRM. Average trips per customer should be tracked every quarter / month so the trend can be identified.  

ANALYSIS #2: Location Analysis

In the previous section, we introduced the concept of a transaction distribution, and we outlined a couple of drill pathways with this metric (i.e., loyalty members vs non-members, loyalty redeemers vs non-redeemers). One additional drill pathway warrants its own section because of the diagnostic precision and organizational engagement it can yield.

Breaking out a transaction distribution at the location/store level leads to deeper engagement from local managers with your loyalty and CRM efforts. This is critical because local managers are an important link in the chain between upper organizational management and the customer. They have an intimate understanding of their store and the environment around it. And, empowering them to spot opportunities and diagnose challenges within their own location will improve the performance of your loyalty program overall.

More importantly, location level analysis serves to enable store-level marketing initiatives, which is one entry-point into personalized marketing.

Analysis #3: Value Segments

Not all loyalty members are created alike, and not all loyalty members will yield your business the same value. Knowing how much value each member brings to your organization can help you tailor your engagement and relationship with them.

Members can be categorized into transactional “value segments”. Knowing how many of your members falls into each segment is powerful in and of itself. However, to truly harness the power of this categorization, you can take it a step further and tag each customer inside your database with their value segment (this is similar to categorizing members as “base, silver, gold.” )

Just as in our previous analyses, this analysis can and should be filtered by location to understand the make-up of particular stores’ membership. Differences between locations can help you pin-point areas of opportunity and success and can help you precisely target the locations that need the most help, saving you resources and cost in the long run.

Once you have these value segments set, it is easy to use them to understand your loyalty membership on an even deeper level. For example, creating metrics by value segment like average ticket, average spend, and total spend will help marketers at the corporate level and managers at the local level better understand customer value. Comparisons between locations can inform pricing and promotional initiatives.

Analysis #4: Value Segment Overlay

One way we go beyond the basic metrics, is to combine research and analytical methods to create insight that isn’t possible using more traditional means of data analysis. Here is one way to do so, using the analytical foundation we’ve set up in our previous sections.

Many companies employ surveys to understand their customers’ motivations, needs, and areas of satisfaction and dissatisfaction (or if they don’t they easily can). The value segmentation we previously described creates a mechanism to draw better insight from the profile / preference information you collect in member surveys.

Overlaying survey data with value segments allows marketers to understand what make their best customers (the whales) loyal, and what aspects of the value proposition they should promote to grow lower value and new customers.

Analysis #5: Occasion Analysis

Ultimately, a main driver of customer behavior is the “occasion.” By this, we simply mean, an environmental set of circumstances that motivates a purchase. For example, one particular family might always visit the same restaurant every Sunday for brunch. The circumstance trigger (Sunday) motivates a family to visit a specific restaurant each and every week.

In many industries, creating new occasions is an important objective. The first step in doing so is understanding when your current transactions are happening. Transactions are time stamped, and it’s easy enough to segment members by day part. Gas stations, convenience stores, movie theaters often think in terms of day parts.

Occasion / day part analysis has many uses. One obvious application is that it can influence staffing and merchandising. Going beyond this, for marketers, understanding if customer value tends to increase because they visit more often for the same occasion (our Sunday brunch example above), or if customers naturally have new occasions as value increases, can help shape targeted marketing messages that drive highly incremental behavior.

Analysis #6: “Activation” Analysis

While attracting new customers is important, what really matters is whether existing customers have a second and third visit. Every loyalty program we have seen tracks enrollment—the simple number of customers that sign up for the program.  But what really matters isn’t just how many people enroll (or new customers we acquire), you also need to know how many of those new customers turn into repeat customers.

Simply put, how well is your loyalty program working for you by creating repeat customers?

An analysis of average number of trips in a new customer or loyalty member’s first 30 and 60 days will help you answer that question. This analysis helps you understand if you are “activating” new customers and building frequency and habit, versus just creating one and outers.

An activation analysis is particularly useful to help program managers test and assess alternative communications plans. Rather than an A/B test the open rate of a particular mailing or promotional response to a particular offer, this approach allows measurement of the impact of alternative communications treatments on the creation of new purchasing habits, which is where the real value lies.

Analysis #7: The “Leaky Bucket” Report

We see it time and time again. A company works hard to acquire 100 new customers, and at the 1-year mark, 90 of those customers are gone. This is the dreaded “leaky bucket.” Companies and loyalty programs often face this problem. They are constantly acquiring new customers and enrolling new members, but active customers / members aren’t really growing that fast. Often nearly as many members/customers are lapsing as new members /customers are being acquired. Two steps forward,1 step back. (Sigh.)

The degree to which you have a leaky bucket problem signals the amount of effort your company should be putting into building a rock-solid customer engagement and retention strategy.

We visualize that leaky bucket problem by doing an analysis of new members a given year, compared to lost members (the number of members who had activity in the prior year but had no activity the current year). This report can also be done on a quarterly basis, or any other timeframe that makes sense for your cadence and business cycle.

Analysis #8: Uptrender/Downtrender Analysis

Last, but not least, a key business metric for retailers, dealers, restaurants, convenience stores, and many other industries is Comp Store analysis. Simply put, a comp store analysis examines whether sales are up or down, year-over-year, for a cohort of customers who were active in both measurement years. This is a powerful metric for retailers because it allows them to quickly see if they are gaining or losing ground with their active customer base.

A similar metric can be calculated on a customer or loyalty member basis. With a slight change to the calculation, you can measure, whether sales/transactions were up, down, or even for customers who were active in the current period and the prior period.

The logic for calculating this metric is straightforward. First, identify all customers who were active in both periods (years, quarters, months), then calculate the difference in sales/transactions between current time period and prior time period.   If the difference is positive the customer is an uptrender, if it is negative he/she is a downtrender.

An uptrender / downtrender analysis is a powerful way to understand the effectiveness of loyalty and CRM efforts. Said another way, are you winning more often than you’re losing? This analysis will tell you.

Go Beyond the Basic Metrics and Make your Loyalty Program Work for You

In this post we presented 8 different analyses that can help you use data science and analytics to build and run a more intelligent loyalty program. This knowledge goes beyond the basic metrics to turn your relationship marketing program into a stronger, dynamic and more meaningful customer experience. The analyses include:

  1. Transaction Distribution
  2. Location Analysis
  3. Value Segments
  4. Value Segment Overlay
  5. Occasion Analysis
  6. Activation Analysis
  7. Leaky Bucket Report
  8. Uptrender/Downtrender Analysis

So many companies don’t realize the value that lies within their customer databases and loyalty programs because they don’t reach beyond the basic metrics .

Want to riff about your metrics and growing your program? Give us a shout.