December 23, 2007

Channels And Facts

As usual, many folks in the catalog/retail industry communicated with me in the past week about the role of a catalog in our modern multichannel world.

Communication sometimes breaks down when talking about online orders that are "matched back" to a prior catalog mailing.

Catalog purists consider these to be "catalog orders".

While I haven't talked about this concept on the blog, I view these orders different than most industry experts do. These orders are a combination of advertising channel (catalog) and physical channel (online).

The matchback algorithm gives credit to a catalog, and the analysis stops.

My thinking evolved considerably over the past twelve months. These days, I'd prefer to run a Multichannel Forensics simulation on each combination of advertising channel and physical channel. It is the interplay between advertising/physical channel combinations that matters. Our industry just doesn't talk about this.

I would code each order for a customer, over the course of two years, into the following set of combinations (not all may be valid for your business, there may be more combinations for your business, use this as a guide):
  • Catalog Key Code, Order Mailed To Company.
  • Catalog Key Code, Order Via Phone/Fax.
  • Catalog Key Code, Order Via Sales Staff.
  • Catalog Key Code, Order Via Website.
  • Catalog Key Code and Paid Search URL, Online Order.
  • Catalog Key Code and Natural Search URL, Online Order.
  • Catalog Key Code and Portal Advertising URL, Online Order.
  • Catalog Key Code and Shopping Comparison URL, Online Order.
  • Catalog Key Code and Affiliate URL, Online Order.
  • Online Order, E-Mail Coded Order.
  • Online Order, Paid Search URL.
  • Online Order, Natural Search URL.
  • Online Order, Portal Advertising URL.
  • Online Order, Shopping Comparison URL.
  • Online Order, Affiliate URL.
  • Online Order, Direct Load or No-Source.
In this example, there are sixteen "channels", if you will. Before any simulation is run (and this requires a healthy simulation ... forecasting sales five years from now in each of these "channels" is not a trivial exercise), I'd run the migration probability table (this step is exceptionally easy to do). The table illustrates how customers who ordered from any of the sixteen "channels" last year behaved this year.

And this is what is really important, folks. For instance, where does a customer who ordered via "Catalog Key Code, Order Via Website" migrate to in the next twelve months? Does a customer who only orders online ever use a catalog key code in the future?

If that customer always utilizes a catalog key code in the future, then you know that the catalog matters as an advertising channel.

However, if this customer migrates to ordering online (without a catalog key code), or starts ordering with online advertising URLs, then the catalog is "losing" this customer to other advertising channels.

There are many readers of this blog who are critical of the content, especially when it comes to catalog marketing. All too often, critiques are based on beliefs, on non-data-related observations.

I'd love it if catalogers ran this type of analysis, and proved that cataloging is vital to the future of the direct marketing industry. Heck, I'll publish your findings as an article on this blog for 1,100 folks to see, findings that prove the essential need for catalog marketing.

I'd be just as happy if catalogers ran this type of analysis, and found out that customers were no longer using the catalog channel like they did in the past.

I'd be happy either way, because catalogers and vendors were using actual facts to defend or refute the craft of cataloging.

No comments:

Post a Comment