When I took a job at Community Coffee as Consumer Direct Manager, my marketing experience was almost completely in the online world. I could run PPC, Organic SEO and affiliate marketing as good, or better than most, but I had a unique challenge. Community Coffee’s B2C division was about 50 years old, and at least half of those customers had no email address. The order management system didn’t even have a customer field for their email address. I knew I had to learn the catalog industry quickly. How was it that I could keep getting those 4-color catalogs from companies I had little history with? How can Gevalia possibly afford to send me so many letters with rich offers of free coffeemakers, plush bathrobes or mugs just to join their coffee club?
Well, there’s nothing I love more than a new challenge, so I embarked on a mission to figure out the mystery. Like a true ‘net guy, I researched online and I bought many books about “database marketing.” One of my favorite books is from someone I’ll call “the godfather of database marketing.” Those in the industry may or may not scoff at that, but Arthur M. Hughes became my godfather and mentor for the upcoming years. The book is called Strategic Database Marketing and the experience I received from the book was incredible.
Who knew there was such a database marketing gospel full of completely virgin vernacular: RFM, LTV, deciles, regressions, Nths, Quintiles and so much more. I realized that I was an acquisition expert, but I had no idea how to retain and remarket effectively to my existing customer base. I was both excited and overwhelmed. Where do I start? The answer for me and for most of you who are still stuck on the word Quintile is to start simple with RFM and LTV.
RFM? Well that stands for Recency, Frequency and Monetary of course. Oh, that’s right, you’re still stuck on the word quintile. Wash that out of your head and focus. The point of RFM is that “People who have bought from you recently are much more likely to respond to a new offer than someone who had made a purchase in the distant past” (Hughes.)
So wouldn’t it make sense to group your customers according to when their last purchase occurred, or Recency? If you were to send a direct mail piece to your customer list, do you think Joe will respond to your offer for 10% off on his next order if he just purchased yesterday? What about Sally who typically purchases every 60 days (her buy cycle) and it’s been 55 days since her last order ? If you group your customers and market to each group, Joe may get a brand building piece or a satisfaction survey, whereas Sally will get a 10% off offer and Tom, who typically has a 30 day buy cycle, and it’s been 60 since he last ordered, will get a richer free shipping offer, or a BOGO (buy one get one free.)
Now that you understand the role of Recency, you can apply the same logic to Frequency, or how many orders each customer has contributed. A customer that has only placed a single order with you will traditionally be worth less than a customer with 2 orders. I’m not saying don’t pursue customers with 1 order, just control your spend as your response rate will be considerably less. So now you can group your customers according to frequency and market to each group differently. For example: customers with 1 order, customers with 2-3 orders, customers with 4-6 orders, customers with 7-10 orders and customers with 11 or more orders.
Customers can also be grouped according to their Monetary contribution, but as Arthur Hughes points out, the monetary segmentation produces less reliable results, and for the most part, I didn’t use this segmentaion in any area other than for customer service so our CSR’s would know if they were speaking to a “good customer.”
Once you have your RFM segmentation complete with a R value, an F value and an M value assigned to each of your customers, it can then be your primary weapon in determining which groups, or “cells” are profitable and which are not. What if you could take a cell and mail to a small sample (yes, an Nth) of the customers in the cell to measure their response rate. If all of the customers within a cell have similar buying characteristics, it would make sense that they would respond similarly to your offer. If your mail test proves that it was an unsuccessful offer, then you’ve saved a lot of money by not mailing to the rest of the customers within the cell. What if your offer is profitable? Well, I’m glad you asked; you then mail to the rest of the cell and reap in the rewards of your brilliant marketing strategy.
Once you build RF or RFM codes for all of your customers, use that power in the online world as well. Target introductory offers to those with few orders and little frequency, or email richer offers to those that are outside their buy cycle. The possibilities for marketing efficiency are endless!
As much as I wanted to drone on about LTV, or Lifetime Value, I’ll have to save that for a later post. You’ve got a lot to read