As marketers, we often use a variety of data in order to segment our prospects and customers for our marketing programs. Geographic and demographic data, product information and customer type data is used most often.
These are important but you can improve the effectiveness of your campaigns and get better marketing results when you include data on actual behavior.
Segmentation means using data to sort your addressable audience into groups with common characteristics that make them likely to be motivated by similar things.
Actions speak louder than words
In high level terms, there are two key ways you can segment your data:
- Using descriptive data such as geographic data (country, address), demographics (age, gender, income) or firmographic data (industry, revenue, size)
- Leveraging observed behavioral data (actions taken by the customer)
- Interactions: website visits, product usage, etc.
- Purchases: the number and value of the orders)
While descriptive data is important, research has shown that including behavioral data gets you higher results. And this is where RFM segmentation comes in.
Marketing programs based on behavioral data have shown to deliver 11x better results than individual campaigns that just rely on descriptive data
The Three Pillars of RFM Segmentation
The recency of the customer activity, in purchases or visits.
Time since last purchase, last website visit or last product engagement.
|Recent: Within 4 months
|Not recent: 4 to 11 months ago
The frequency of the transactions or website visits.
Total number of orders/visits or the average time between orders/visits.
|Frequent: 3 times or more
|Not frequent: Less than 3 times
The comparative value of the purchases by the customer.
The Total or Average Order Value (AOV) the customer spends with you.
|High Value: Above AOV
|Low Value: At/Below AOV
How to get started with RFM segmentation
Step 1: Build your RFM behavioral segmentation model
Determine what recent and frequent means for you and what the average order value is. You can use the one we describe in the table above with two tiers or you can define your own.
Step 2: Divide your customer list across the tiered groups
Using your marketing automation platform, sales platform or Excel, divide your customer list across the tiers you defined.
Step 3: Add other relevant information to segment
Though important, using only RFM to target is likely not enough. Add persona, geography, demography and product information where relevant.
Step 4: Determine which audiences to target
Based on your goals and the characteristics of the group, select the customers you want to include in your campaign. The table below provides some example goals per group.
Step 5: Determine the tactic and craft the tailored content
When you have set the goal for the groups you want to target, determine what action you need to take. See the table below for an overview of actions that fit each particular group. Focus on the behavioral characteristic of the selected groups to determine the best tactic and create content that allows you to be more relevant and more effective.
Where to go from here?
To further refine and optimize your segmentation approach there are a number of approaches your can take. First of all, the overview we presented here only has a limited number of tiers — for Recency we only set ‘recent’ and ‘not recent’. You may want to experiment with more tiers to define your audience.
Secondly, your product offerings and go-to-market may be different per geography and you may be targeting different personas. You will then want to combine RFM behavioral segmentation with other information such as geography and demographics to get better campaign results.
Thirdly, if you are keen to predict future behavior, RFM has its limitations. There are advanced customer segmentation techniques based on predictive analytics techniques which are more accurate at predicting future customer behavior.
If you have any questions or would like to know more about driving marketing results, go-to-market strategies and levering marketing technology, let us know. We would love to hear from you!