Lessons From the Data Brokers


Despite its name and author, the new report entitled “Data Brokers: A Call for Transparency and Accountability” from the Federal Trade Commission makes a fascinating read. First of all, what’s a data broker? I believe this label originated within the government to describe aggregators of information about consumers, excluding credit bureaus. As you might expect, this definition includes an eclectic mix of companies. For example, the nine randomly selected companies asked to provide background information for this report are: Acxiom, Corelogic, Datalogix, eBureau, ID Analytics, Intelius, PeekYou, Rapleaf and Recorded Future. The operating scale of these businesses is impressive: one reported that it maintains over 700 billion aggregated data elements, another holds information on 1.4 billion consumer transactions, and still another adds 3 billion new records monthly to its database. And the money is substantial too, with just these nine companies generating over $427 million annually from the sale of consumer data.

So are there lessons for data publishers in the activities of these large providers of personal data? Absolutely. Here’s what I see:

Let’s start with creativity. There seems little doubt these consumer data companies are a number of years ahead of most B2B companies in extracting maximum value from their data. They have developed market segmentation systems, ratings and scores, and powerful analytical tools. They understand the value of historical data to discern patterns and trends, and have rolled out lucrative new products based on data others might discard. Also, these companies truly understand and apply the concept of inferential data. Someone with a pickup truck and a fishing license can be categorized as an outdoor enthusiast, for example.

Perhaps most powerfully, these companies are active in helping marketers to bridge the online-offline divide. They are routinely matching online website registration data to the deep offline data they collect, creating much richer audience profiles. These companies will even embed some of their data in tracking cookies for ad targeting purposes. And of course there is intense marketer interest in understanding online-offline buying behavior.

The segments of the business are interesting as well. Data brokers, in the view of the FTC, break into three types: those who primarily sell marketing data, those who sell risk management data and those who sell people search products (the “background check” and “find anyone online” products that are proliferating these days). Again, we see a progressive industry. It has morphed from the obvious marketing application for its data to risk management. Risk management in the B2C world largely means using data to pre-screen purchases. For example, a risk management product might alert an online vendor that the buyer has had merchandise shipped to an unusually large number of different addresses, a potential indication of fraud. So what’s largely the same dataset has been spun into an entirely new market, and very successfully we might add: for the nine companies in the FTC study, risk management revenues are rapidly approaching marketing revenues. Are there untapped B2B opportunities in risk management? I believe there are.

We also see that with the people search products, the industry has morphed from selling its data to businesses, to selling its data to consumers. This is a sweet pivot that many data publishers aspire to, because with little more than a different user interface, you’ve got a product of interest to a new and vast market.

Another intriguing insight from the report is that none of the companies surveyed obtained all their data from the original source. All were licensing substantial portions of their data from each other. And that makes sense because it allows these companies to move faster, and not have to develop the large staffs necessary to be expert in so many different data sources. Why build your own capability to pull in home ownership data from the source when you can license it from someone who gathers it as a primary business activity? And these licensing agreements, not surprisingly, have grown complex, with some even containing clauses preventing “reverse engineering” of licensed data elements as a way to keep other data brokers from getting too clever with licensed data.

What this report offers is an inside peek into the operations of some very savvy, large and successful consumer data providers, and there you’ll see the future of the business data industry as well.

Comment