It has been my distinct pleasure to work for three companies who have, arguably more than any others in the world, applied the theories of statistics towards solving real-life business problems. Today I thought I would discuss one example of applied science from each of these companies and then explain how that knowledge is similarly being used to advance our technology vision at Inuvo.
In the mid 1990s, a little known San Diego Company was struggling to find its place in California’s tech boom. CEO Bob North, a former TRW executive, along with what can only be described as a collection of the smartest people I have ever been associated with, came across a unique observation while working on a project for one of their bank customers. They noticed a set of patterns in the credit card transaction purchase data. These patterns where not random – when someone typically buys gas, then groceries, then something at Target, and all of a sudden the next transaction is an expensive diamond, it just might be fraud. HNC Software went on to commercialize a Neural Network based analytical-fraud technology within a software solution which is now universally used around the world and known as Falcon. Next time you get that call from your credit card company, you’ll know what prompted their action.
The parallels are strikingly similar within Inuvo’s business. In our pay-per-click (PPC) business for example, the ability for individuals to perpetrate fraud by manually or technologically clicking on links without any intent to buy is unfortunately quite common. Since advertisers are all too willing to pay for clicks, this business model demands that the service provider (like Inuvo) protect the advertiser from this potential fraud threat.
Much in the same way HNC Software predicts fraud on credit card transactions, Inuvo predicts fraud on click transactions. On the cost-per-action (CPA) side of our businesses, where advertisers are willing to pay based on the creation of a lead, fraud is also a potential problem. Again here, Inuvo is applying the same proven principles that protect hundreds of millions of credit-cards. By opening the kimono for the advertiser on every single transaction, the Inuvo Platform has the ability to assign each transaction a fraud score, which can help those advertisers improve return on investment (ROI) by eliminating this potential fraud threat.
The power of analytics goes well beyond fraud in terms of its application. In the early 70s, two clever engineers, Bill Fair and Earl Isaac revolutionized the credit systems of the United States and even the World. They, like HNC, developed an algorithm that was based on an individual’s credit history and predicted with great accuracy the relative risk of lending to the consumer. The FICO™ score went on to become the standard mechanism for assessing risk in the U.S. Again here, the parallels to Inuvo are striking. In the FICO score case, the historical actions of consumers—how they paid off their historical loans— was a powerful predictor of their ability to service their current loans.
This is also true in marketing online. How an individual has historically responded to advertisements, supplemented by various information about their current interests, lifestage or geographic location, can be used very effectively to predict what advertisement to present today. The same is true in the FICO example — replace the outcome decision from “what loan to offer” to “what product to offer”, and you have an analytical match.
Inuvo is implementing numerous data-driven analytical techniques to help advertisers and publishers on the PPC and CPA sides of our business improve the response rates for their advertising offers. While many consumers might say that these techniques make them nervous, as a marketer, I have always felt that a well-targeted offer meeting a current consumer need is really a win-win.
The third example applicable to our Inuvo business is innovation that occurred at Acxiom®. Out of the small town of Conway, Arkansas, a marketing services powerhouse emerged. Charles Morgan built Acxiom from nothing into a $1.3 billion annual revenue business and Acxiom, like FICO and HNC, leveraged data and analytics to give them a competitive advantage. At its core, applied analytics is really the science of codifying the meanings hidden within vast quantities of data. In the 1990s, Acxiom innovated on two fronts that have parallels for us at Inuvo.
First, Acxiom invented a universal way to identify consumers. Many organizations today have consumer relationships that span many years and many departments. A car loan, a credit card and perhaps a mortgage, along with the history of those relationships, probably resides in multiple databases across the enterprise. So in a typical organization, there is a high probability the company has a record of many individual relationships with, say, John Smith. To the organization, all these John Smiths are different people, but are they? An analytic approach to understanding the underlying data is more likely to reveal that a John Smith from Arkansas is the same as a John W Smith who used to live in Florida and Jonathan Smith who went to school in Texas. Acxiom created a way to make these associations, for everyone in the U.S.
Second, Acxiom created a consumer segmentation system that grouped various demographic characteristics at the household level so products and services could be better targeted based on lifestyle and lifestage. Essentially, they took all the demographic information they had and applied it at the household level so marketers could understand the differences between lifestage and income, along with many other characteristics. For example, on my street, like I’m sure many of you, we have individuals who recently had children, we have a couple who are grandparents, others who recently retired and ourselves, who have two kids in college. Clearly, each of us on this street is at a different lifestage and, as a result, will likely respond very differently to various marketing offers – the Howes no longer need diapers.
The combination of these two innovations in science allowed advertisers to 1) ensure they grouped all the relationships they had with a consumer together and 2) ensure they properly understood that consumer so advertisements could be targeted appropriately. Inuvo is approaching online advertising in much the same way Acxiom and others had approached offline advertising.
The online problem has many of the same challenges addressed in the 70s and 80s. An individual visits a web site but never identifies himself /herself. How do we know whether or not they are an individual Inuvo has interacted with in the past? Further, suppose we could, at a minimum, realize that we had a former relationship with this consumer—how do we know what they might be interested in? Inuvo is implementing technology today within the Inuvo Platform, within our PPC market and in our owned and operated websites that will for online marketing, assure analytically, that every consumer interaction aligns the offer with the highest probability of response.
The thing that made Acxiom, HNC and FICO successful were their innovations in data and analytics. Inuvo similarly recognizes that true differentiation online will likely come not from the software platforms themselves, but rather the intelligence built into those platforms. To succeed, one needs both, and Inuvo is and will continue to make investments that are aligned with this philosophy of data and analytics driven advertising decisions—we’ve even hired some of our friends from HNC, FICO and Acxiom to help us.
