Big Data, Big Hype

The individual responsible for "enterprise growth" at American Express was recently quoted in this newspaper, saying about mobile that the "data is more valuable than the payment transaction." We're told this preference for data contradicts Amex "tradition." No, it contradicts Amex experience.

In the past, Amex repeatedly invested in activities that were supposed to harvest value from data. It acquired control of First Data more than 30 years ago. Later it spun this off.

As I recall, Amex also used to have a joint venture aimed at mining marketing data about individual consumers — but that's disappeared. In 2009, Amex set up American Express Business Insights to sell aggregate data and analytics. Does that generate big earnings?

Now, Amex executives talk about getting consent from individual consumers to sell their transaction data because "that data is tremendously valuable to marketers." Well, the company had plenty of opportunity to do this over the course of 50 years. Why is data from cellphones more valuable data?

Which is more lucrative, selling transaction data to marketers — or handling the actual payment transactions? Is it really the data, as Amex now claims?

Visa has a capitalization of roughly $100 billion, based on revenue from payments. Just a smidgen of Visa's value derives from selling data and analytics. Contrast that with Experian, the credit bureau. Experian doesn't do payments, but has lots of data and some analytics on almost all consumers. It's worth one-tenth as much as Visa. Competitor Equifax, which sells both payment and consumer income data, has a market value only half of Experian's.

Banks are vast repositories of consumer data. Capital One Financial, almost a monopolist in subprime bank cards, has unique data on tens of millions of people. Its value would be reduced by over 90% if it only had data and wasn't lending people money. It used to be in the cellphone business and its executives never expressed regret about shutting it down. Maybe cellphones and their data won't be pivotal to banking after all.

Unlike Visa, Capital One could easily use its data to sell consumers merchandise. Remember the product inserts that used to come with credit card statements? With all the transaction data, does Capital One make money marketing nonfinancial products to customers? Not much. Same for PayPal.

Grab the other end of the telescope. Groupon has a business model and underlying it a market targeting system that appears to generate value. Among the endless offers they send me, none have shown the slightest cognizance of my buying interests, or even of my location. Looks like success doesn't hinge on precise targeting of offers. After all, electronic junk mail costs only a tiny fraction of what older channels and media cost.

Store couponing is far bigger. After decades, many stores finally learned precisely what items their steady customers buy. Nirvana! But I don't see them targeting these customers with offers. For instance, if I buy Dole canned peaches, the store could offer me a bargain on their private-label peaches. But they don't, and neither does Walmart.

Big money is being laid — by Amex and others — on the proposition that the teenager who always has his cell in hand anyway will, at checkout, choose between PayPal and his prepaid card by pressing a button on the phone. Assuming we ever solve the formidable security problems.

That's great, except that every attempt at a "virtual wallet" on the personal computer or tablet has failed miserably. And the wallet yields far less consumer convenience at check out with a cellphone than on the PC — particularly now that lots of websites won't retain card information. Type that again and again. Shades of the Amex Blue card reader.