Getting to The Core Of Catalog Marketing Advice

I received an email recently from the CEO of a company that, like Datamann, is a supplier to the catalog industry.  He was commenting on one of my recent blog postings, and offered an observation about a problem that he saw within the industry.

He mentioned that many online media outlets are reporting on companies that use Artificial Intelligence (A.I.) in their quest to find new customers and correctly target existing ones. He stated that in most of these articles the term A.I. is misused. The companies profiled are instead using “just predictive models (typically neural networks) and not true generalized ‘artificial intelligence’ at all. But claiming the use of AI is all the rage among Marketing Tech vendors (and the online media reporting on their activities) and causing massive confusion among the marketers I speak with.”

I knew exactly what he was meant. At a company I was previously associated with, every new service or product that was introduced – no matter how mundane – was promoted as having “advanced analytics”.   So we’ve progressed from advanced analytics to artificial intelligence, and both turn out to be an exaggeration.

This vendor went on to describe many of the innovative services and products that his company provides their clients.   Some of them sounded pretty cool – even cutting edge. But he wasn’t trying to impress me or sell me anything. Instead he wanted to make a point that so many of us have to contend with. Some of his products are really meant to make a mailer’s job more efficient and easier. But what is the client’s typical response? “Cool – but now what should I do? Can’t the system just act on these insights for me?”

He bemoaned the death of fundamental marketing skills among most catalog mailers, stating that “the Staples ‘Easy Button’ for marketing has become fully realized and it’s doing far more harm than good at many companies today.”

We all see it. When I polled mailers and marketers for what they wanted to learn at our seminar next week (yes, it is now less than two weeks away), the marketers I contacted almost universally asked for “the five most important metrics I need to run my business.”  When I explain to mailers that those metrics are different for everyone, they lose patience. They don’t want to hear that – they want me to skip to the “best” metrics and reports that everyone else uses.

I can’t blame mailers for not having the patience to understand that their business – as is everyone else’s – is unique. I’ve mentioned before (perhaps ad nauseam) that there has been a talent drain in the catalog industry. Maybe a more accurate statement is to say there just aren’t many staff left at many catalogs. It’s not that the individuals left lack talent – it is simply a fact that in just about every catalog company, the jobs of five people eight years ago are now being done by one or two people. The reasons that this has happened are endless.

It Is Always A Step Backwards

When staff is downsized, it is almost always a step backwards. What happens when someone is downsized or leaves and you are left to do your job and the other guy’s job? You cheat, and take short cuts to get your job done. Maybe cheating is not the right word, because it implies that what you did was ethically improper. That’s not the case. But you need to find a way to get more done, so you look for the easiest route. One of those routes is to outsource your marketing and circulation planning.

There are plenty of consultants in the industry that will do circulation planning and/or modeling for clients. The vast majority of these folks are ex-mailers. They know their stuff, and generally they are all good at what they do.  I perform the circulation planning for a handful of Datamann clients. My methods are no better or worse than those of other consultants. We each have our little quirks that make how we do it different from the other guy – but no one has a secret sauce, beyond experience.

But here is what bothers me. I view circulation planning as being a core function of the catalog. That’s because that was the role I filled as a cataloger. Of course I want to think of my job being “core” to the success of the company. I have always felt it was a mistake for catalog companies to “farm out” their circulation planning. To me it is a relatively easy exercise. But when you have someone else do it, you become that much further removed from the business. Once you start letting someone else plan the circulation, after a few seasons you start to skip looking at the reports that consultants like me provide that show how the last mailing performed, and how the next one is planned. You start to lose touch with how your customer is performing.

This phenomenon of losing contact with your customer is even more pronounced when you have someone doing your circulation planning via modeling. At least with RFM, if you wanted to, you could see how each customer segment is performing. But modeling requires a huge leap of faith. You are often mailing huge swaths of your customer file in very large segments, simply defined as Segment 1, Segment 2, etc. Yes, you can see a difference in response between the segments, but can you tell which portions of your customers are not responding?  Do you even take the time to ask the modelers, or do you simply assume that they are doing the best job that can be done?

It’s one thing to hire a consultant like Kevin Hillstrom or Frank Oliver to come in and provide you with an assessment of what you are doing. Maybe even have them build you a model. But, consultants like Kevin, Frank and me are always willing to teach you, the mailer, how to do what we just did for you, so you can replicate it and carry on the process when we are finished with our assignment. Rarely do mailers want to do that. They want that “Easy Button”.

I’m not knocking companies that do hire outside modelers. There are many of you that are big enough to justify modeling, but not big enough to hire your own in-house statistician to do it.  The problem is that it is not that much of stretch to go from the “easy button” of modeling to being confused by discussions of artificial intelligence to think that your model should be able to “learn by itself” what to do next.

I don’t see this as an issue with every mailer. People that have been “around for a while” know what is, and what is not, possible. It’s the younger professional, mostly from the ecommerce side of the business that suddenly find themselves also responsible for the catalog that think there should be an algorithm for everything. “Facebook can determine what kinds of people are most desirable to view my ad, why can’t your model just figure out the 2% of the people that are going to respond, and mail to them”.   They of course understand that Facebook’s response is not going to be 100%, but can’t understand why the postal model should not be attaining 100% response.

In my opinion, there is no “easy button” to catalog marketing. You have to be involved. You can hire others for their expertise, but don’t become too reliant on them. If you put too much of your business on “autopilot”, you will lose touch you’re your customer, and what they are doing. Your company will become an example of the old joke about the guy that jumped from a 10 story building, and people kept hearing him say as he passed each floor “Okay so far”.

If you are not already signed up for emails from this blog, click here.

by Bill LaPierre

VP – Business Intelligence and Analytics

Datamann – 800-451-4263 x235

blapierre@datamann.com