In the June issue of Catalog Success magazine, catalog veteran and consultant John Lenser — in "RFM Turned Upside Down - How online ordering is changing the age-old circ plan formula" — made one of the most astonishing claims I've come across in a long time. Consider the following:
What do you suppose would happen if you mailed a print catalog to 1x-buyers who placed their order on the web? Do you think they'd reorder at higher, lower or the same rate as the same web shoppers who did NOT receive a catalog?
What about multi (2x+) web shoppers (again, catalog vs no catalog)?
What about web shoppers who hadn't placed an order in well over a year?
Internet buyers often behave quite a bit differently from their catalog counterparts. So I would have guessed that mailing the catalog would have little to no effect on the web channel's reorder rate. As it turns out (and this isn't in just one isolated test case, either) the catalog mailing significantly depressed the reorder rate under all three of the above scenarios! Consistently and dramatically, too. In short, he observed that web shoppers who received a catalog reordered at a substantially lower rate than those who did not. Here's a few excerpts from his article:
"In a recent test we conducted with a client, we created two panels of 50,000 customers each, both groups acquired on the marketer’s e-commerce site within the past three months — neither had been mailed a catalog.
One panel was mailed a 48-page catalog; the other was not mailed. Based on a later matchback to the mail files, the one-time buyers not mailed the catalog responded at 3.7 percent, while those mailed the catalog responded at 2.1 percent. Of the two-time-plus buyers, those not mailed a catalog responded at 8.5 percent; those mailed responded at 3.6 percent.
Panels also were created for older buyers. For buyers who previously had purchased more than 13 months ago, those not mailed responded at 3.4 percent; those mailed responded at 2.8 percent. This isn’t a one-time test. We’ve repeated this test with several clients and seen similar results."
This is fascinating and worthy of careful consideration given that one might actually be losing money - and not merely wasting it - by sending catalogs to internet-driven shoppers at all. In an attempt to get a better handle on the implications of this finding, I wrote to John with a few questions …
Lloyd: What niche/type of product(s) did the catalogs offer (e.g. apparel, consumer electronics, home furnishings, etc.)?
John: This data came from two pure plays that had NEVER mailed anything until we started mailing a catalog for each of them. That is why the tests could be so large. The companies the data came from was home decor and travel supplies.
Lloyd: What time of year was the test catalog dropped?
John: The tests were done several times throughout the year.
Lloyd: And are these seasonal businesses?
John: No
Lloyd: Any idea of the half life of the catalogs?
John: Three weeks
Lloyd: How does the average order size for the catalog compare to that on the web?
John: Catalog order is slightly larger.
Lloyd: Was a repeat mailing ever made? (If so, were the results similar?)
John: Yes (and yes).
Lloyd: Did you test mailing the catalog with an incentive not available to a routine web shopper (e.g. free shipping on any order)?
John: NO.
Lloyd: I'm assuming, regardless, that internet shoppers do NOT receive an offer that is in any way better than that for a catalog buyer.
John: True. What is important to note with both these companies is that since they are e-commerce companies, there are many e-mail promotions going to all customers. So, the catalog is competing with numerous email offers. Our conclusion is that any response to emails suppresses response to mailed offers. So, there are multiple reasons why response to a catalog would be less.
Lloyd: But isn't this a separate issue entirely? The issue, I thought, was the realization that mailing a catalog could (in itself) depress reorder rate. That, in other words, customers who were mailed a catalog sometimes subsequently reorder at a lower rate than an equivalent segment who were not sent a catalog. Whether or not email campaigns were taking place simultaneously (and presumably identical for both the test and control group) is irrevelant, no? Or did I miss something?
John: No, the data did show reduced response of those receiving the catalog. The only explaination for this is that the customer felt betrayed that their e-commerce company had mailed them print material. Frankly, I do not trust this explanation and we will also be running the test again. Mailers need to do their own tests and also segment their house file by how the customer originated.
Lloyd: I seriously doubt that "feeling betrayed" is the only possible explanation. As is usually the case, there are probably several factors at work here.
John: Interestingly, other tests have shown that those who unsubscribe to emails have lower response to mailed pieces.
Lloyd: Isn't this to be expected, though? Those who don't want to receive emails (i.e. communications from the company) would more likely than not feel the same way about direct mail.
John: The real point is that the new paradigm for response analysis is the "difference" in response from those mailed vs.. those not mailed. Control groups are critical.
Okay, it looks like a pretty good bet that this is reliable report. Lenser is a highly respected consulting firm and very unlikely to have made any sophomoric, let alone fatal, mistakes in either designing or carrying out the tests lead to these results. In particular, that the results were consistent and quite similar across disparate recency and frequency groups tells me that there's something significant and consistently reproducible here. The question is what?
MY TAKE: As mentioned earlier, I would have guessed that mailing a catalog to web-channel shoppers would have little to no effect on future reorder rate. That a significantly negative behavioral effect was observed came as quite a shock and belies strong psychological undercurrents at work. As is nearly always the case when analyzing group behavior — and most other phenomena, come to think of it – it is not one but a constellation of factors that are responsible. I’d like to suggest three reasons why mailing a catalog might significantly depress reorder rates (listed in descending order their perceived impact):
- The catalog "offended" them This is the most obvious explanation. Many people hate to receive unsolicited mailings. For some, it’s about “the trees.” For others it’s just the “junk mail” label itself. Either way, once they’ve gotten that bad taste in their mouths, their reluctance to experience that taste again (whether via catalog or website) may be sufficient to account for the reduction in their likelihood to reorder.
- The catalog “soured” their opinion of the company Perhaps simply by receiving the catalog, the shopper no longer viewed the company as “a cool place {read website} to shop,” but instead is now seen as another one of those catalog deals, and, therefore, not worth bothering with.
- The catalog “satisfied” their urge to shop Web shoppers tend to be impulsive and the ease with which they can place an order (it’s just a few clicks if they’re a repeat buyer) makes it dangerously easy for them to give into that “I want it.. NOW!” voice in their head. Not so with a catalog where far more than a couple of clicks is required to place an order. Instant gratification isn’t possible. The buyer must first load and log into the company’s website (when browsing, on the other hand, they’re “already there”). Here’s the point though … the buyer may have satisfied that urge to shop merely by being exposed to the offer/catalog, particularly if they’ve thumbed through it. In the process of doing so they may have even felt inclined to order something, but couldn’t give in to that urge so easily because no shopping cart was at hand. After all, they’re staring at a paper catalog so the probability of them placing a new order — given the extra effort this would entail — is far less than were they able to have clicked and checked out as they’re accustomed to doing. In short, web shoppers are impulsive buyers who like to click things into their shopping cart. Hand them a catalog and they don’t act because the path of least resistance (their beloved shopping cart) isn’t available. But they might feel as though they just spent some time shopping and won’t feel the need to do so again with this company for awhile.
So … do you agree with this assessment? And are there other factors worth considering?
// lm
August 23rd, 2007 at 4:13 am
Other possible reason: people didn’t see what they wanted in the catalog and dropped the company from their list of places to consider buying from, not realizing the Web site would include additional products not shown in the catalog.
September 8th, 2007 at 12:59 pm
I’ve seen this phenomenon a number of times. The effect is normally fairly small, and normally affects only a fairly small proportion of the population, but it is present to some extent in many different types of direct contact.
In fact, I have worked with one client (who I can’t name, unfortunately) who uses control group and uplift modeling very systematically, and who claims that in almost every campaign he runs, the last one or two deciles (ranked by uplift, i.e. difference in treated purchase rate minus control purchase rate) show negative effects.
I think there are a set of reasons, and don’t really find it all that surprising. I think your first two reasons are big. A customer’s opinion of a company or a product can easily be changed (including for the worse) if a piece of marketing alters their image or perception of it. For example, if a customer thinks a particular mobile handset would be perfect for business, and receives a communication that shows kids fooling around swapping videos with it, that may make the handset suddenly seem less appropriate. Or more prosaically, if a mailing features a celebrity whom the client happens to hate, that could put him or her off.
Interestingly, there is one area here negative effects are even more prevalent, and that’s in retention activity, where attempted retention efforts often actually trigger defection. I’ve written about that at http://stochasticsolutions.com/retention.html.
January 11th, 2008 at 1:19 pm
I believe that the main force causing decreased response rates discussed above results from the email campaigns that preceeded or ran concurrently with the catalog mailed. The email satisfied the demand for the offer for this customer group.
CMC has noticed increased response rates when it blasted an email asking customers to “look out” for the upcoming catalog in their mail box. Also, I’d love to know how many sku’s these Lenser companies had. Currently, internet browsing does not come close to replicating the experience of flipping the pages of a paper base catalog.
March 7th, 2008 at 12:10 pm
I would like to apologize for a couple of things before continuing with this thread…
First, I was preoccupied with a number of other things (in particular, selling my other software company so I could focus exclusively on Deltalytics) and completely neglected this blog for several months. I’m back on it, though, with renewed energy and a higher sense of purpose. Second, my blog’s spam filter didn’t behave as expected and these comments were held up in a queue I didn’t notice until now. Gee, I hate when that happens!
My responses to these much appreciated earlier comments follows –
(1) To David Raab who said “Other possible reason: people didn’t see what they wanted in the catalog and dropped the company from their list of places to consider buying from, not realizing the Web site would include additional products not shown in the catalog.”
– This is an interesting - and I think perfectly valid - twist on my original thought that .. ‘The catalog “soured” their opinion of the company Perhaps simply by receiving the catalog, the shopper no longer viewed the company as “a cool place {read website} to shop,” but instead is now seen as another one of those catalog deals, and, therefore, not worth bothering with.’ However, I’d be willing to bet that the catalog, at least in in this case, fairly represented the company’s product mix if not completely. // I’ll check with John Lenser to see and will report back.
(2) Nick Radcliffe provided an insightful perspective on negative retention effects and notes that he has observed this phenomenon many times in his consulting practice.
– Nick included a link to his excellent “Optimal Retention Strategies” article [ see http://stochasticsolutions.com/retention.html ] wherein he describes the mechanisms by which attempts to positively engage the customer can have the opposite effect, viz. defection. Sometimes, in fact, it’s best to leave happy customers alone (so as not to annoy them); likewise the unhappy ones (reminding them of their displeasure can elicit an undesirable behavioral reaction, e.g. realizing and deciding that they never want do business with you again). It’s a tougher and more delicate balancing act than is typically acknowledged. In short - and I believe Nick would agree - the risk of promotions depressing reorder rates (including, if not especially, retention efforts) can and should be mitigated whenever practical; but can never be eliminated entirely.
(3) Martin Davis said: “I believe that the main force causing decreased response rates discussed above results from the email campaigns that preceeded or ran concurrently with the catalog mailed. The email satisfied the demand for the offer for this customer group. CMC has noticed increased response rates when it blasted an email asking customers to “look out” for the upcoming catalog in their mail box. Also, I’d love to know how many sku’s these Lenser companies had. Currently, internet browsing does not come close to replicating the experience of flipping the pages of a paper base catalog.”
– I’m not sure I completely follow [a] even if Martin’s assumption - that email campaigns ran prior or current to the catalog drop - is correct. Even if the email(s) had some behavioral impact, this observation begs the question of why the catalog-mailed segment would have reordered at such a dramatically lower rate. Indeed, no what matter else may have been going on during the catalog group, we must presume that no bias existed between the two test panels. Here again is Lenser’s summary to refresh our memories ..
Even if these two panels were treated differently (a sophomoric mistake I don’t believe the Lenser people are likely to have made), we’re still left with the question of why the catalog-mailed group reordered at a significantly lower rate? Which brings us back to square one, viz. the puzzling observation that receiving the catalog “put off” these customer somehow. Back to Martin’s comment … I’m not surprised that a heads-up email advising customers to look for an upcoming catalog in the mailbox had a positive effect. This can create a positive anticipation and associated void in the customer’s mind (i.e. something to look forward to) which is fulfilled and relieved when the catalog arrives. All it takes is for some statistically meaningful number of customers to open the catalog, rather than tossing it immediately into the trash, to increase response rate.