Jackson Barrett looks at his mail a little differently than the rest of us. The flood of glossy letters — what the industry calls direct mail marketing, and what the rest of us call junk mail — isn’t a random deluge as it might appear to the untrained eye. The odds that your neighbor got the same credit card offer that you did are in fact very low. Jackson knows this more than anybody, since his company, Resonant Analytics, is in the business of mining the rich veins of consumer information that companies now have access to for useful insights into consumer behavior. A smart company will use that data to send offers that are not junk at all, but which the recipient might actually find tempting — and that’s what Jackson looks for when he opens his mailbox.
The fact is that the businesses of the world would like to get to know you a little better. Are you the kind of person who buys a car once every 10 years, or who leases a new car every six months? GM sure would find that information useful. If you rack up a credit card bill, do you pay it in full every month, or do you carry a balance? Citibank would like to know that before sending you a credit card offer. How about if you use Photoshop — are you a professional photographer, a graphic designer for a business, or an art student with a pirated copy? Adobe software could tailor its marketing E-mails to you based on that information.
Barrett founded Resonant Analytics four years ago with the aim of helping companies answer those questions. In June the home-based business moved into a new home on Vaughn Drive, where five employees work. Resonant, and its sister companies owned by Wiedemann & Lomasney, employ 50 people all around the country.
Resonant Analytics doesn’t collect all that data, it just analyzes it. One reason for Resonant’s quick growth is that companies have more access to more kinds of data than ever before, and more kinds of companies are looking for ways to use it.
Barrett grew up in Detroit, where his father was a pipefitter for a Ford truck factory. Barrett says his dad didn’t want his son working in a factory like him, so he encouraged him to be an academic. His grandfather, a coal miner in Scranton, Pennsylvania, had made sure his own son got a good job in a factory instead of working in a coal mine. (Barrett says he would be happy if his own three children went into data mining like him.)
He earned an undergraduate degree at Eastern Michigan and worked in Chicago for a few years before getting his doctorate in computational chemistry at the University of New Hampshire and beginning post-doc work at the Naval Research Lab in Washington. His goal was to become a university professor, but an unexpected opportunity in business sidetracked him.
Computational chemistry is the science of building mathematical models to describe chemical reactions. At the time Barrett earned his doctorate, Wall Street companies were recruiting theoretical physicists and statisticians to predict market trends. “Analytics and data mining is all about developing algorithms to identify patterns,” he says. “There are a lot of similarities with computational chemistry. A lot of it centers around optimization of pattern recognition to understand not chemical system behavior, but human behavior.”
Barrett found his credentials were suddenly in demand by companies that had a lot of money to throw around. He could also see that universities were starting programs to turn out just the kind of specialists that finance companies were recruiting — Carnegie Mellon had just launched a computational finance degree program — so the demand for academics was destined to be short-lived.
Barrett decided to jump on the bandwagon before he lost the chance. In 1996, he joined a startup in Maryland called Neuristics that was applying advanced statistical techniques to credit scoring. The company was developing an algorithm, using neural network-based artificial intelligence, to identify patterns and better predict risk in the sub-prime credit market. Although borrowers classified as sub-prime had a high default rate at 15 percent, that meant that 85 percent of them would pay their bills.
Neuristic used advanced math to analyze the subprime market and further divide the population, and sold the resulting credit scores to lenders. One of the company’s backers was Louis Renieri, the creator of the mortgage-backed security.
After three years, Barrett left Neuristics to work at Grey Global Group, a New York marketing firm. His first job there was to take the techniques for analyzing credit ratings he had learned and apply them to the job of optimizing credit card offers on behalf of Chase. Barrett worked for the consulting arm of the marketing company, which is now known as G2 Knowledge Consulting.
But by the mid-2000s, the credit card companies were becoming increasingly sophisticated themselves, and less in need of hiring outside consultants to figure out their credit ratings. “Chase has 50 to 100 statisticians building models,” Barrett says. “We were this little group trying to beat our heads against the internal guys.”
That was when G2KC discovered a whole new market for data analysis services. It turned out that while the credit industry had been crunching the numbers for years, other gigantic companies had access to just as much data, but weren’t using it as much. Unilever, a massive global maker of consumer products, had been gathering consumer data but didn’t know how to use it for marketing purposes. It hired G2KC to work on its Slim-Fast product line. Another new client was Adobe, the software company that makes Flash, Photoshop, Acrobat, and other programs.
Barrett took all the information about Adobe’s various programs — how many people clicked on the ads they were running, how valuable a customer was to the company, and other such information, and used it to target marketing efforts. Adobe had previously had a policy of sending frequent e-mails to customers, but Barrett found that the marketing could be more effective (and less annoying) by being less frequent and more tailored to the customer’s particular needs.
In 2004, Barrett took over managing Adobe’s business for Grey, which he continued until 2011, when he founded Resonant with the help of George Wiedemann and Kenneth Lomasney. Along with Resonant, W&L owns marketing firms Drum and UMarketing.
Jackson decided to locate Resonant’s headquarters in Princeton, where he has lived since 2002, for several reasons. He originally moved to the area when his wife got a job at Bristol-Myers Squibb. “I had the choice of a New York location and Princeton,” he says. “I felt that this was a good location to attract the type of talent that we would need.”
Today Resonant counts Adobe among its clients. It has also worked for the pharmaceutical company Novartis, the global hotel chain Iberostar, the video game maker Electronic Arts, and Time-Warner Cable.
Resonant specializes in analyzing data to solve business problems, sometimes but not always related to marketing. Barrett believes it’s important to start off with a goal in mind and organize the data that will help solve the problem, rather than starting with data and seeing what can be learned from it. “What happens when you start off on the data side is that you end up with an ocean of data,” he says. “The problem is understanding what’s important and using that information to generate insights.”
For example, one of Resonant’s clients is PHH Corporation, a transportation company that manages vehicle fleets. PHH had a network of mechanics all around the country to service its cars. Dealing with many “mom-and-pop” vendors was expensive, but increased convenience for its clients. There was a happy medium between having enough vendors to keep clients happy, but not so many that the cost became overwhelming. Resonant looked at an ocean of data, including service records for thousands of vehicles, and found that the company could save money in the long run by dropping some of the vendors from the network while maintaining customer satisfaction.
The PHH case is the kind of optimization problem that data analysis is great at solving, and which is being used more and more in the business world. Resonant, which also has an office in Chicago, is expanding rapidly. Jackson recently hired two people to work at the Princeton office, and is planning to open a location in San Jose for its Silicon Valley clients.
As a businessman as well as a data analyst, Jackson says he is used to communicating between a company’s statistical team and its management. The numbers, after all, are just numbers until someone gleans meaning from them. “We’re about bridging the gap between data and decision-making,” he says.