Applying the RFM loyalty model to customer purchases at the category level provides actionable insights on how retailers can increase basket size and decrease churn.
Many retailers use the Recency, Frequency, Monetary model (RFM) to analyze and compare customer loyalty levels. However, analyzing loyalty at store level provides only a partial view of the overall loyalty picture. Using big data to calculate an adaptive RFM model across category levels, reveals basket splitting and churn, and consequently can be used to generate marketing programs that increase basket diversity and prevent churn.
Source: ciValue.
The secret, as explained in a previous blog post, is to analyze RFM at different levels of the category tree—at department, category, subcategory levels, and even at brand level. This analysis reveals rich insights that can identify low-hanging fruit and quick wins to get customers to deepen their loyalty and expand their basket.
While there are numerous ways to leverage category-level RFM analysis, we present 4 especially useful insights based on performance analysis:
1. Splitting basket
If your goal is to convince customers to become one-stop shoppers, doing so requires an understanding of how customers split their shopping baskets between your store and others.
RFM at category level reveals which customers have a low loyalty score in a core category. Perhaps those customers buy their core products elsewhere. Perhaps, counterintuitively, targeting customers with offers on core categories in which they are not loyal can help increase their basket by incentivizing a shift of their purchases to your store in those specific categories.
2. Churn at category level
A loyal shopper has reduced the frequency of his frozen food purchases recently. Past behavior suggests that he buys frozen food regularly. While this may indicate a lifestyle switch, it is more likely that he or she now buys frozen products elsewhere. For many customers some categories are more important than others, and losing them on those categories can mean losing them entirely.
Tracking category-level churn can alert you to take preemptive steps to save lapsing or churning customers. Timely, targeted, personalized offers can win the customer back.
3. Changing shopping behavior
As I’ve mentioned in the previous section, many customers change their lifestyle and consequently shopping behavior over time. Wine enthusiasts may “graduate” from buying wine at the grocery store to patronizing specialist stores or ordering online. Young parents eventually switch from baby formula to baby food as their infants get older.
Timely identification of behavior changes while difficult unlocks a world of marketing opportunities. Even discerning wine enthusiasts will find it hard to resist specials on premium wines and sending the right wine offers can actually increase their loyalty. Anticipating parents’ upcoming needs with timely offers may increase overall spend.
4. Expanding categories
Another great hidden revenue opportunity is customers who are Platinum loyalty customers at one category, but only Bronze at store level. Take a drugstore for example. A customer may be coming in to buy only personal care products, never buy any beauty or household goods, which make them Bronze at store level.
This is a lost opportunity, as these customers already set foot in the store. Therefore, targeted offers may help them expand into more categories and departments, and so expand their tier.
Conclusion
Customers are different. But using big data can establish recurring and similar patterns among a diverse group of customers. While using the RFM model at store level reveals our best and worst customers, it provides a very partial picture. Using RFM at category level can identify the necessary preemptive steps to strengthen loyalty, increase customer lifetime value and reduce churn.
This post is a part of a series of blog posts covering loyalty, retention and personalization.
To learn more about the RFM model, subscribe to our blog.
