Reinventing Retail Origination: The Role of Big Data

The following is an excerpt from Chapter 13 of Volume I of The Mortgage Professional's Handbook:


Garth Graham, Managing Director, STRATMOR Group

We might do much better as an industry if we followed the approach used by candidates during the most recent U.S. presidential election, which happens to be the same approach used by leading consumer product marketing companies. Like them, we should be leveraging big data.

Avoiding a discussion of which candidate any of us thinks should have won the election of 2012, no one can deny that the use of big data analytics by the various candidates played a huge role in the outcome, with one inescapable conclusion being that the campaign that made the most use of those tools won the office. The Obama Administration, the incumbent in that contest, made masterful use of superior analytical methods to uncover the constituents known as swing voters, those who had not previously committed to a candidate and were not even considered likely to vote. By winning these wildcard voters, Obama paved a path to a second term in the White House.

The result shocked many who were not as well versed in big data marketing tools as the Obama camp. A good example of the contrast in approaches can be seen in the way the Romney campaign focused on traditional mass marketing techniques and general polls to set its course, while the Obama campaign used new techniques to target specific voters and get them out to vote. It was an effort few thought credible, but the results were truly impressive.

The reason it worked was because the campaign didn't try to win over every American voter, which would have led to a great deal of wasted effort. Instead, they sorted through the consumer data that was available to them, identified the voters who were still on the fence, and made them the focus of their marketing efforts. Obama won because his team fought the battle they could win, the fight for those voters in the middle who were living in the states that mattered (from an electoral college perspective), and getting them to vote.

Mortgage marketers of the future will need to do the same thing.


The reason big data—the sort of data the Obama campaign used so brilliantly— is so compelling, and will be a much better solution for tomorrow’s bankers, is that the younger generations of American consumers {particularly the Millennials, but so too with the others} are blazing digital trails that marketers can easily follow.

There are already information and data analytics products available that can aggregate information for all three hundred million consumers in the United States of America and sort it down to the address level. These tools will tell us when the homeowner last took out a mortgage on a piece of real estate, which is an excellent predictor of when they may go back to the closing table for another. We can also learn the approximate interest rate of their current mortgage, which is a fairly decent predictor of whether or not they may be interested in refinancing. People in the data set who do not have a mortgage are candidates for first time homeownership. Knowing how old they are, which we can easily determine, will tell us when they may be ready to purchase that first home. And all of these questions can be answered before we even begin to apply the advanced analytics that are available for use on big data.

With a big data strategy, the home finance industry can use the information we have available to us— countless data elements—and sort through the hundreds of millions of people in the United States, sorting and analyzing the data in an effort to predict which individuals will likely need a mortgage so that we can target our marketing directly to them and begin building a relationship. Far from eliminating the LO from the equation, this approach allows our front-line salespeople to reach out and build relationships with those borrowers most likely to be receptive; our own swing voters, if you will.

No longer will our industry waste resources marketing our products to people who don’t need them. We will use big data to narrow down our prospects to the ten percent who are in the “buy zone” and focus on them. Instead of paying a hundred basis points to entice loan officers to go out into the community and search for these people, bankers in the future will pay a fraction of that to send analysts and their computer models into the data to seek them out. Now, what some readers are likely to say is “yes, but mortgages and real estate are local.” That may be true, in which case the smart mortgage bankers will spend money on local sales teams tasked with building those specific relationships that will yield new loans. Gone will be the days of chasing a million real estate salespeople or spending millions of dollars on mass marketing for a tiny return.

Read the rest of this chapter in The Mortgage Professional's Handbook!

Garth Graham heads STRATMOR Group’s marketing strategy and execution practice, which focuses on lead generation and lead management methods and practices primarily for the consumer direct and retail mortgage origination channels. Graham has over 25 years of experience in sales and marketing, ranging from Fortune 500 companies to successful startups, including management of two of the most successful e-commerce platforms.

As an executive with ABN-AMRO, Graham was part of the executive group and author of the 5-year strategic plan that led to the successful $9 billion sale of the mortgage company to Citigroup in March of 2007 which closed before the market downturn.

Graham was a founder and key member of the executive management team at, and led the company through its dramatic expansion in the late 90s and successful IPO in 2000, before its ultimate sale to ABN AMRO. originated the first mortgage through the Internet in 1997 and was the original platform for the Quicken and Microsoft mortgage initiatives.

Graham also worked for nine years at Chase Manhattan Mortgage, where he developed the corporate and affinity marketing programs. His regular columns in National Mortgage News and American Banker are widely read. Graham studied Economics and Computer Science at the University of Virginia and has completed Six Sigma Champion and Green Belt Training. He lives in Fort Lauderdale, FL with his wife Joan and teen-aged daughter Nancy.