There are still some new tricks
A few weeks ago, I wrote about the death of the catalog media, like Multichannel Merchant magazine. In preparation for doing so, I went out to the loft in my garage where I keep old catalogs and old copies of Catalog Age (predecessor to Multichannel Merchant) to get one to use a photo of in the posting.
I looked through a few old issues of Catalog Age from the late 1980s and early 1990s. There were articles that went into great detail explaining the benefits of NCOA, using an 800 number, talk time in the call center, mailing to Canada (that topic never seems to go away), and the best way to allocate multis on a remail. It was a nostalgic trip back to the topics that were so important when I first started in cataloging 30 years ago.
Now where are we? All of these marketing wonders of years past are now taken for granted, or are misunderstood. We no longer debate the pitfalls of desktop publishing, whether to include a FAX number on the order form (most of you don’t even have an order form), and your list broker, who you counted on to negotiate your list deals, has retired. We can’t even count on the likes of Hershell Gordon Lewis and Don Libey to warn us of our excesses and mistakes.
There are still many neat tricks that you can do with a catalog to drive response. But, just like I’m sure a lot of great blacksmithing secrets went to the grave when cars replaced horse and buggies, so too are the many secrets of cataloging going to disappear.
So let me give you a new one for your circulation planning.
Most of you segment your file by RFM, which is fine and adequate for many of you. You further segment by channel – typically by whether a buyer purchased by mail or phone (typically labeled “catalog buyers”), or whether they ordered via your website (typically labeled “internet buyers”). You may also break out your retail store buyers and Amazon-only buyers (which I recommend).
But, it’s the web/internet buyers that can be further segmented to improve your response rate when mailing a catalog. During matchback, those “web buyers” that matched a mailed catalog can be identified as “catalog from web buyers”. These buyers are not true web buyers – they still used/needed the catalog to respond. They will perform better in future mailings than the pure web buyers (customers acquired through SEO, PPC, email, etc.) with whom they were previously grouped.
When Datamann identifies these customers for clients, we keep these “catalog from web” buyers separate from those catalog buyers that ordered over the phone, partly because they perform at a slightly lower response than mail/phone buyers, but also because in the future, they may be the customer to whom you can mail a smaller catalog aimed at driving these buyers to the web.
In general terms, these would be the difference in magnitude in response you would experience between these groups of buyers, when comparing equal RFM segments:
Mail/phone buyers = 3 per cent
Catalog from web buyers = 2.75 per cent to 2.50 per cent
Remaining (pure) web buyers = 1.75 per cent to 1.25 per cent
If you model your names for mailing, your models probably already take this performance by source into consideration. If you use traditional RFM for your mailings, this adds an additional set of segmentation for buyers, and requires an additional step (identification of the names through matchback), but the added value in the difference in response is worth the effort.
Finally, by identifying those web names that in theory, needed no catalog to respond, you have identified those names to which you potentially do not need to mail a catalog, and have potentially reduced your overall circulation expense.
There are still some good tricks to employ to drive response with your catalog. This one is simple to identify, easy to implement, and very profitable. Go to it.