into Profits with a Spreadsheet
The Guide to Maximizing Customer Marketing ROI
 |
 |
 |
 |
The Web Retailing Example - Drilling Down Newsletter # 32: April 2003Drilling Down - Turning Customer Get the Drilling Down Book! Also check out: The Marketer's Common Sense Guide to E-Metrics - 22 benchmarks to understand the major trends, key opportunities, and hidden hazards your web logs uncover. I wrote this manual with Bryan Eisenberg of Future Now, the visitor conversion specialists. To find out more about this topic: click here Prior Newsletters: In This Issue: Hi again folks, Jim Novo here. This month we've skipped the usual "best of" Customer Marketing article links and put more time into the Recency case and the question from a fellow Driller, who is about to launch a version of the Recency-based discount program the owner of IMissAsia.com tested last month! Amazing coincidence, eh? My favorite topic lately is this one: why are companies so afraid to call an end to the customer LifeTime? If you can't peg the LifeTime, you can't measure LifeTime Value, and if you can't measure LifeTime Value, you leave the majority of CRM ROI off the table. Sound interesting? If you want to find out how to attack this issue and bring home the ROI bacon using simple customer models, check out my searchCRM interview and webcast: WebCast (on-demand, ignore "start time") OK, let's do some Drillin'!
Recency: The Web Retailing Example Recall last month, the owner of IMissAsia.com discovered how to lower discount costs while increasing response rate by using the Recency metric to modify offers. Testing the 91-120 day Recency bucket (last purchase 91-120 days ago), the owner found it was more profitable to offer a 15% discount than the standard 10% discount to customers in this bucket. You can see the test results here. The owner had learned a lot in the past few months, from understanding how to define an "active" customer to predicting the likelihood of a customer to respond to a promotion; from improving the profitability of newsletter promotions to actually being able to predict the profitability in advance based on the Recency buckets. Despite all this, the owner of IMissAsia.com still had not answered the original question that started it all: why was the response rate to newsletter promotions falling while sales remain flat month to month? The owner started packing boxes, thinking about how Recency had proved to be so significant a factor in customer behavior. While packing, the owned recognized the names of best customers, and also the names of some of the new buyers. Why were these new buyer names familiar? While pondering this question, the owner realized where the names had appeared before - on the daily new subscriber list to the newsletter. Then an idea struck the owner like a bolt of lighting: since Recency of purchase predicted the likelihood to purchase again, is it possible that Recency of subscription could predict the likelihood to make the first purchase? Again with this Recency thing? The owner's head was now spinning, this was too much, too fast. It was a struggle just to go into the subscription records. Figuring out when Newsletter Responders had joined the newsletter seemed like an insurmountable obstacle. The owner was sweating, looking out at the spreadsheet with barely one eye open, murmuring "too much, too fast." The owner's hand trembled on the keyboard, as the last table join came into view. And there it was. Looking at Only Buyers, they subscribed to the Newsletter How Many Days Before they made a First Purchase?
The owner gasped aloud. How long ago someone subscribed to
the newsletter predicted whether they would respond to the newsletter
with a purchase! And how long ago someone responded to the
newsletter and It's that whole "cycle" thing showing up again - customers seem to pass through "stages" and these stages are predictable based on the previous behavior of the customer - as long as you know how to measure and track the behavior. If you are using the wrong metrics, the owner thinks, you simply can't see or take advantage of these cycles. Well, fellow Drillers, we know what the owner of IMissAsia.com is talking about, don't we? It's the Customer LifeCycle, the most powerful tool you could possibly have to increase the ROI of a marketing campaign and increase the profitability of a customer. If you understand what a customer is likely to do before they do it, then you can plan for and customize offerings and campaigns appropriately. Marketing based on the Customer LifeCycle is "event-driven" marketing. You use LifeCycle metrics like Recency to mark or define customer events, and then spend only when you have to, and always at the point of maximum impact. This creates extremely high ROI marketing and service programs, because the targeting mirrors customer behavior. Most of the buying activity was coming from people who Recently joined the newsletter. Since the number of new people per month joining the newsletter was flat, sales are flat month to month. But as the newsletter list keeps growing each month, this "new blood" - new subscribers who are likely to respond and buy - becomes a smaller and smaller percentage of the total newsletter list. So sales remain flat as response rate is falling. Newsletter response riddle solved. But how can you increase profits knowing this? Well, if you offer discounts to people who have just signed up for the newsletter - the people already most likely to purchase - then you probably are giving away margin you don't have to give away. Save your discount budget for those less likely to purchase, customers who are deeper into the LifeCycle, the ones who are "less Recent." The owner continues to ponder these cycles, with the likelihood of a customer to move to the next step in the sequence at any point in time ruled by the Recency of the customer for the prior step. The owner had read some stuff about CRM and wondered if this "sequence" was what all that CRM noise was about. Maybe the owner should see if the rep who called yesterday "stands in awe" of IMissAssia.com. Perhaps, the owner thinks, I could learn "how to keep customers front and center" without hearing "Five Hundred It Is" from this rep. No, "CRM" would have to wait. It now appears increasing new subscribers to the newsletter is the most important thing that can be done to increase sales, because Recent newsletter subscribers are the most likely to buy. And after all these discoveries about Recency, the owner wondered if Recency could have something to offer on the newsletter subscription challenge. What do you think? We'll get the answers next month, as only the data knows for sure. To read the next installment of Recency: The Web Retailing Example,
click here. Q: Hi Jim! I purchased the Drilling Down book and just love it. A: Well, thanks for your kind words! Always good to hear people are "getting it." Q: We're a lab supply company. I'm getting ready to do a web promo based on Recency ratings (we're starting with just doing Recency ratings until we get our feet wet). What we plan on doing is to rank our web customers based on Recency and put them into quintiles as you suggest. We will then offer a sliding discount promo and will then measure the ROR and ROI versus the customer Recency score. A: I take it you mean Recency of purchase, not
visit...since you're a customer now, I'll The crux of the matter is usually 1x buyers - they can be very Recent, but they are less likely to respond than multi-buyers. It would be instructional (if you can) to do a basic Frequency select by creating two groups - 1x buyers and everybody else. Score each for Recency separately and then do your sliding scale promo as planned to each group. You will probably find it works very well for multi-buyers, but results by Recency are a bit random for 1x buyers, with perhaps product of first purchase being more predictive of the discount needed to optimize profits. Do you follow what I mean? If you divide the groups initially, it should be pretty easy to see these patterns. If you rank them all together on Recency only, your ranking will be skewed with 1x buyers and might not be optimal. This approach is discussed under the "Hurdle Rate" idea in the book pg 94 - 101. Q: I'm concerned about the situation where several people within the same company get drastically different e-mail promotions because one person has a "5" Recency rating and another has a "2" Recency rating. Do you have any ideas on this potential problem? Of course, if a customer called and complained, we would certainly give them the higher discount, but I don't want to have to do this on a wide-scale basis. On the other hand, maybe most customers within the same company won't notice and I'm worrying about nothing... A: Well, you've got your thinking cap on, and it is a minor concern. Two areas: 1. My experience is this: it is much more of a concern in the call center than with the customer. It is critical all customer service agents understand what is going on and understand the policy of granting the discount a "buddy" is offered. If you keep track of who gets what discount, you can audit on the back end how much "noise" gets into the system. It is typically quite low, maybe 1 - 3% of total response (not of total mailing), and with e-mail being such a personal thing, it's even lower. E-mail is different than dropping catalogs with a "15% off any order over $500" dotwack on the cover into the same place you mail catalogs with no offer. If the customer is treated "matter of factly," they don't care and the issue disappears. If they are treated to a "hell no" experience, then they start talking about it with other people in the lab and then you get them all on alert to share info and all of them order using the best offer. 2. It really is more of a concern when you have duplicate customer records and the same customer gets two different offers. So to the extent possible, try to de-dupe the list. But it really isn't a big deal. I get two catalogs from Dell Computer every few weeks and they both have different offers, one better than the other. I buy from the one with the better offer - but I still buy. Then I become more Recent with the one catalog and the offers get better in the other. But I still buy - from Dell. Ultimately, that's the mission, and if you consider the "company" as a customer, this kind of stuff evens out in the end. After all - they could buy their equipment elsewhere, and as long as you are making money on the promotion, I wouldn't get too wrapped up in the dynamics of offer profitability at the individual buyer level, especially when you are just starting out. Get the "gross" picture moving in a positive direction, then go back and see if maybe you can tweak for specific situations. In other words, this stuff works on the "aggregate" level. There will always be individual customers who you lose out on and beat you on the offer, it always happens. But in the aggregate, you make a lot more money. Just make sure your customer service people understand the game - which does NOT mean giving away the best discount, it means giving the customer the discount they ask for, as long as they have the correct promo code. Let me know if I have explained this well enough for you to go ahead and execute! Jim
------------------------------- Any comments on the newsletter (it's too long, too short, topic suggestions, etc.) please send them right along to me, along with any other questions on customer Valuation, Retention, Loyalty, and Defection here. 'Til next time, keep Drilling Down! - Jim Novo Copyright 2003, The Drilling Down Project by Jim Novo. All rights reserved. You are free to use material from this newsletter in whole or in part as long as you include complete attribution, including live web site link and/or e-mail link. Please tell me where & when the material will appear. |
 |
|
|
Slow connection?
Same content, less graphics, think Jakob Nielsen in Arial - Go
to faster Contact me
(Jim Novo) for questions or problems with anything on this web site.
| |
|
Get the book with Free scoring software at Booklocker.com Find Out Specifically What is in the Book Learn Customer Marketing Concepts and Metrics (site article list)
|
|
|