You see, these days, the classic e-commerce buyer doesn't need many catalogs. This isn't 2003 - catalogs don't drive online volume like they used to. Instead, we mis-attribute online demand to catalogs - often badly. I've been observing this since founding MineThatData in 2007.
We know from most mail/holdout tests that, for catalogers, +/- half of online demand will happen anyway, even if catalogs are discontinued.
Furthermore, there is interaction between catalogs and email campaigns. Catalogs typically push a modest amount of demand to email campaigns (i.e. the more you mail, the better email campaigns perform). Conversely, email campaigns typically cannibalize a small portion of catalog demand (i.e. the more emails you send, the more discount/promotional messages you offer, re-directing catalog demand to email promotions).
As a result, we have to evaluate email contacts and catalog contacts simultaneously. If we increase email contacts, we reduce catalog demand. In other words, we must run a profit and loss statement on every combination of annual catalogs and annual email campaigns, after accounting for the interaction between each channel. The combination that yields the most annual profit is the combination that should be assigned to a customer.
More on this topic tomorrow. And we'll throw paid search into the mix!