JRosenfield Article

RELATIONSHIP MARKETING, 2004

James R. Rosenfield

March 2004

The term "relationship marketing" has been around for probably 20 years, and the concept forever. It should be a no-brainer: relationships are good, so cultivate them, especially with good customers! But when marketers talk about it, they're like a person who wakes up each morning, sees the sun, and says "What is that big bright thing?"

Case in point: a long article in the January 25, 2004 New York Times , about IBM. "Last summer, CEO Samuel Palmisano asked IBM employees to share their thoughts online about what the company's future values should be. The results were distilled to three: customer relationships, innovation and trust.Corporate customers don't so much buy technology products, since the technology changes so fast, as invest in a relationship with a trusted supplier."

All of this is undoubtedly true, but even mentioning it is akin to observing that humans ambulate on their legs. When it comes to customer relationships, we either belabor the obvious, or descend into techno-babble. "Consider the case of FinnAir.IBM sold mainframes to it for decades and has run its data center operations since 2001. Last spring, IBM began going further: its services and research lab started working with FinnAir on a project to use mathematical modeling and optimization algorithms to try to increase customer loyalty, reduce marketing costs and improve response rates among members of its frequent-flier program - the 10 percent of its travelers who account for more than half of the airline's revenues.

"The program involves processing huge amounts of customer data and using sophisticated math techniques developed by IBM researchers to predict how the frequent flyers behave. FinnAir is pleased with an initial project involving half of its frequent fliers.the technology has reduced marketing costs by more than 20 percent

and improved response rates by up to 10 percent."

Now, for all I know IBM is indeed performing miracles for FinnAir, but I've been down this route a lot, and here are some questions that occur to me are:

--What's this little project costing FinnAir? Big Blue does not work for small green.

--Half of FinnAir's frequent fliers is not a large number, given Finland 's small population and out of the way location. Is the rollout potential enough to justify the costs?

--How poorly was FinnAir doing this stuff in the past, i.e. was IBM miraculously squeezing better results out of an already brilliant program, or was FinnAir on thin ice? I know lots of people who can dependably walk into most companies, reduce marketing costs by 20 percent, and raise response rates by 10 percent, at much lower cost, I suspect, than a giant corporation would charge.

--What does "improved response rates by up to 10 percent" mean, anyway? Did maybe one little tiny segment get that kind of lift? I don't know, of course, but why otherwise would they say "up to?" And what was the offer? If it was free stuff, or something really soft, who cares anyway?

I don't want to pick on either IBM or FinnAir, but the article is a good example of both the painfully obvious and techno-babble. Not to mention a smattering of pseudo science:

"to predict how frequent flyers will behave." No one can predict anything, that's playing God. The best we can do is assigning certain behaviors various likelihoods of occurring and recurring.

It sounds like IBM and FinnAir did a direct mail program. Direct mail can be consummately effective in building relationships with current customers. You can save money and improve results a whole lot, though, without hiring IBM to write "optimization algorithms" for you.

For example, you can always remove costs from a direct mail package. I have seldom seen one that cannot be more cheaply produced without any sacrifice in net results. And you can always improve response from 10 to 100 basis points (the lift depends on lots of factors, to be sure) by doing 25 to 50 ostensibly small things better.

An instance of this?

--Outside of outer envelope, Option 1: name/address show through window on right, envelope copy is underneath name/address.

--Outside of outer envelope, Option 2: name/address show through window on left, envelope copy is above name/address.

Option 2 will bump up your response rate by a couple of basis points. Reasons: People look for their names first.we read from top to bottom, meaning the eye will pick up the envelope copy right away as it involuntarily travels downwards looking for the name.and we read from left to right, meaning that communication is quicker and more natural than if it's on the right.

But that's for another article. My point is, you can improve results without using IBM.

Back to relationship marketing.

A wonderful development over the past several years has been the diminution of the impossibly tiresome buzz-phrase "Customer Relationship Marketing (CRM)." Maybe it's used in back-alleys somewhere in the world, but I sure don't hear it any more.

Another really good development has been the increasing skills on the part of the best on-line marketers, including Amazon.

But Amazon also shows some of the inherent shortcomings of relationship marketing, when it's algorithmatized (did I make up a word?) on the Web.

Amazon makes me offers based on previous purchases. There are two weaknesses here:

• Sometimes I purchase a gift that has nothing to do with my own interests. As a result, I start getting offers for irrelevant products, which of course reduces the warm fuzzy personal feeling that Amazon is trying to produce.

• I buy most of my books these days from Amazon, but what's lacking is the large visual field for browsing that one has in a bookstore. Because Amazon makes me offers based on past purchase patterns, they cannot take advantage of any curiosity I might have for something new and different.

That's why I chuckle a bit when I read about IBM helping FinnAir predict future behavior of customers. You can only do this in a very limited, time-constrained way, unless the customer becomes an active participant in the relationship. Yes, maybe I've been traveling from Helsinki to Paris every month for two years, but what if I've now been promoted and I won't be traveling so much. I'll be much more influential in corporate travel decisions, therefore even more important to FinnAir than I was in the past, but you can't find that out from an algorithm. You need to talk to people. That's where the relationship is. And it's the lack of dialogue that causes so many problems when companies attempt to do relationship marketing.

What guidelines might you consider when contemplating relationship marketing for 2004?

-- Control . The customer is in control of the relationship. You are not. Customers in many categories will not be monogamous, but that doesn't mean they're disloyal. In these cases, your objective is to maximize share of customer.

Control also refers to customer service. Good customer service is the great corrective when things go wrong, and customers feel they've lost control. The most loyal customers are ones with problems expeditiously solved through good customer service. Everyone knows this, but not many companies do anything about it. Customer service continues to be a source of frustration in most categories, in most countries.

-- Contact. How often can you contact customers by mail, phone, and email before they begin to hate you? And now that we are in the world of email, you can do this cheaply. General guideline: The most intrusive the contact, the more customers will hate you. We used to worry about too much direct mail. But in 2004, direct mail is the most benign of contact media, easily to discard with minimum effort. Telemarketing and email are much more intrusive.

Hint: An occasional "love letter" (customer appreciation communication) works wonders, and will frequently increase sales for you.

Another aspect of contact: make sure that customers can reach you through all possible channels, mail, phone, fax, email, retail. Maximize channels, always maximize channels. And 24/7 is now taken for granted. If you don't offer it, you're stuck somewhere in the last century.

-- Complexity. Customers crave simplicity, but we keep complicating things for them. Technology companies produce too many features, and the poor consumer gets overloaded. (What sadistic misanthrope writes those cursed instruction manuals?)

Credit card issuers, responding to perceived competitive threats, add complications to what should be the world's simplest financial product, and customers get turned off.

Make it simple, keep it simple. Complexity stresses people out, and there's far too much stress already. Customers don't want relationships with stress-inducing companies.

-- Change . People want stability, but we keep changing things. An American academic a few years back coined the phrase "cognitive lock-in." Customers get used to things, become disoriented when things change, and as a result defect. A supermarket that reorganizes its aisles always suffers from customer attrition. Does this mean never reorganize your aisles? No, do it, but warn people in advance, explain why you're doing it, make people feel comfortable, have a lot of in-store guidance, and reward customers for learning the new arrangement.

-- Choice. Customers want choice, but only a little choice. Too much choice creates purchase paralysis ("I can't cope with 50 kinds of toothpaste") and buyer's remorse ("I should have bought toothpaste brand X and not brand Y.") Sales go down in categories with too many line extensions. Everyone suffers, including the poor consumer who runs for cover. (There's an excellent new book on this subject, The Choice Paradox , by Barry Schwartz.)

Simplicity, stability, and the human touch: That's relationship marketing. The larger the business, the more difficult this becomes. The supposed solution: Technology. This is one of those things that were supposed to have happened by 1997, and we're still waiting for it. There are hints of progress, but there's still a long, long way to go.