By Murthy Nukala
Founder and CEO – Adchemy

Just a few weeks ago, as Microsoft Chairman Bill Gates testified on Capitol Hill about the state of technology and the need to relax visa standards for foreign technologists, I started to think about the state of technology in our industry.  With Internet advertising evolving every day, I’ve begun to develop this ongoing theory that the technology behind lead generation, particularly when it comes to mortgage leads, is on the verge of something really big.  I’m very curious to know if others agree.

From my perspective, and to borrow a very popular phrase, we’re ready for Lead Generation 2.0. Version 1.0, which served up a mix of bad creative and spotty measurement, is over.  We all learned that lead generation companies can’t just provide leads in volume and then pray they convert.  Particularly today, as the universe of mortgage companies continues to shrink and those that still have online advertising budgets spend less than they once did, all of us are being called upon to explain and ensure the quality of the leads we provide.

The million dollar question, then, is how do you ensure quality leads?  I believe the answer is for mortgage companies to leverage scientific and mathematic principles and create data-driven processes that measure the effectiveness of advertising (including creative and online forms) and then make decisions, or allow systems to make decisions, based on the data.  In my opinion, this type of thinking will drive Lead Generation 2.0.

Let me say, up front, that I’m a mathematician by training, so I have a bias in favor of this type of rationale.  Nevertheless, below are two of the scientific and mathematic principles we can employ that I think can have a tremendously positive impact on lead generation.

Experimental Design – at its core, experimental design is about setting up a rigorous experimental structure that yields the most accurate data.  A lot of mortgage companies study data sets, but the number of them that do it the right way is probably very small.  Some organizations try to determine lead quality and conversion ratio by looking at lead flow from the same time of day (i.e. 1 p.m to 2 p.m.) over a seven day period, others over 30 days.  Some look at the type of lead (i.e. purchase or refinance) and compare that with individual banker experience to get a read on conversation percentage.  This is all well and good, but I suspect that many of these experiments are flawed from the start.  In their quest to act, many businesses forget about control groups, randomization and replications in their studies and are therefore making decisions based on inaccurate data.  Effective experimental design can yield important findings that will aid in the delivery of quality leads.

Statistical Machine Learning – a subset of artificial intelligence, statistical machine learning is the process of programming computers to dissect massive amounts of data and effectively “learn” over time so that better processes can be put in place.  Think about how much time mortgage marketers spend trying to figure out why people respond to one type of advertising campaign over another.  Companies have entire teams of business analysts studying data all day, every day.  Yet we have the ability to adopt technology in which forms and creative change based on the raw data itself rather than “supposed” patterns, guesses, or the hunch of the person reviewing the data.  Statistical machine learning, which can effect changes to online ads in real time, holds enormous promise for our industry.

I understand that talk about “experimental design” and “statistical machine learning” might be overwhelming because the concepts are difficult to implement and the thought of hiring an academic to join your organization might seem a little strange.  However, I also know that science and math is starting to become cool again, and that means we all have access to a wide selection of wonderfully talented young people who can help us improve our businesses and the products we provide.

As reported in a recent Computerworld article, according to the Computing Research Association (the group that follows year-over-year enrollment and graduate trends at 170 Ph.D–granting institutions) enrollment in computer science programs, which declined after the dot-com bust, may be leveling off.  In addition, the Bureau of Labor Statistics projects good growth in IT jobs up to 2016, which should keep science and math majors excited about IT.

As I think about the future of lead generation, I see great power in our community’s use of science and math.  I also know there are a lot of people like me, trained in these disciplines, who are excited to make their mark in the online world.  As an industry, we should leverage the knowledge base that these people have and tap their expertise so that we effectively deliver on the promise that all of us make to our clients – to deliver high quality, convertible leads.

.

.

.