Jim answers questions from fellow Drillers
(More questions with answers here, Work Overview here, Index of concepts here)
Topic Overview
Hi again folks, Jim Novo here.
Yea, I know, so much talk about digital … but does this stuff work for offline businesses? It sure does, in fact, these models were originally developed for offline – way before online was even a thing. But because of the ease of data collection, they tend to work even better online! How about for a natural healing center or an accounting practice? Sure thing! Let’s do the Drillin’ …
Q: I stumbled across your Web site some time ago and have been a regular visitor since. I find your information very useful. You will be pleased to know that I purchased your book (Drilling Down) and have just finished going through it. It all sounds so easy! Your explanations and examples were wonderful and easy to understand.
A: Well, thanks for the kind words. Would you mind if I used the paragraph above as a testimonial on my web site?
Q: Now I will attempt to put it all into practice for two businesses – a Natural Healing Centre (massage, natural medicine etc.), and an Accounting practice.
A: The healing centre is a pretty straight-up situation; should work very well for them just as described in the book. The accountant, as a service business with a built-in “forced” cycle (the tax year), a little more complex. More on this below.
Q: I have 2 questions though, if I can.
A: Sure! The two questions below are related, so I will answer them as one. Only one to a customer! Just kidding…
Q1: Neither business has a Web Site, so a visit to the workplace, usually means a purchase. I was intending to have R = last visit, and F = visits over past 12 months. Will this work?
Q2: Should I put a timeframe on F? The way I see it, if I don’t, F will continue to grow for each customer as long as they are a customer. Whereas if I put a timeframe it will give a better picture of behaviour patterns.
A: The RF behavior scoring model described in the book was developed offline first, so yes, it works very well offline using visits to a store, or deposits at a bank, or for that matter, to predict the likelihood of someone to commit a crime! The likelihood of any human behavior to occur again can be predicted by past Recency and Frequency. The very first studies of this effect: it was used to predict the likelihood of a man to stand when a woman entered the room! And it worked. Goes all the way back to Pavlov and those drooling dogs of his.
Putting a time frame on F is a more advanced application of the RF idea; usually you would only do this after you proved to yourself that customers who have not visited in over a year were not worth scoring. This may very well be the case for many businesses. It will indeed give you a more focused picture of behavior but may also eliminate desirable data on customers with last visit > 12 months.
Remember, the RF scheme is a ranking, comparing customers to each other. So even though the raw number of visits (F) continues to grow as long as they are a customer, the ranking will always be a 5, 4, 3, 2, or 1 as you are comparing customers to each other. The customer with the very most visits, even if there are 1000’s, will have a rank of 5, and the customer with the least visits will have a rank of 1, no matter how long a time period you are measuring This is the benefit of using a “relative” rather than “absolute” system; it “self-adjusts” to any kind of business because it’s based on comparing customers to each other , not to fixed external benchmarks.
So bottom line – if it was me, I’d score all of them first, then score just past 12 months, and test your marketing to see if you get a better result with one or the other. Unless of course you are already sure (and you may be) that customers who have not been customers for over a year are not worth marketing to. As I said, for many businesses, this is true.
Now, with the accounting business, you have “interference” in terms of behavior. Very strong external forces – the tax year, monthly financial statements – dramatically impact customer behavior. I don’t know what kind of business it is (are customers businesses or consumers? do they engage in non-year end tax business?) but you have to consider these forces when looking to predict behavior.
A specialized version of Recency – called Latency – is often more appropriate in an environment where there are powerful external forces like mandated cycles. Latency is about “how long it has been,” usually relative to a fixed date or fixed length of time.
For example, if someone has their year-end taxes done every year for 5 years in February and always makes an appointment by February 15th, and then the next year has not called by February 25th, the customer in “Latent” or their Latency has exceeded the norm for the customer. This tardiness is a signal something may be wrong, and the customer is in fact lost.
Can you see how Latency is more important than Recency for this business? So what if you have a bunch of customers who are Latent (and probably in danger of defection), which ones would you concentrate on? The most valuable ones, the high “F” customers probably. So you can set up an “LF” rather than “RF” type score and still rank customers by how Latent and how Frequent they are. The more Latent they are, the less likely they are to respond or be “recaptured” by any marketing effort.
It is much easier to ring up the guy in the above example on Feb 25th and perhaps get the business before he defects than it is to ring him up a month (or a year!) later and ask for the business – he has probably already switched accountants, right?
If you haven’t seen these articles on the site, more on behavioral scoring in a service biz:
Utilities, Telecom, Insurance – Behavioral and LifeCycle Profiling in Service Businesses
More on Latency (this was covered in the newsletter, but here it is “all together” for ease of reading): Trip Wire Marketing- Tracking Behavioral Change
Hope the above answered your question! Make sure to let me know if I can use your words above as a testimonial on my web site (Jim’s note: she did). Any more questions, feel free to ask.
Get the book at Booklocker.com
Find Out Specifically What is in the Book
Learn Customer Marketing Concepts and Metrics (site article list)
Download the first 9 chapters of the Drilling Down book: PDF