Genalytics unveiled a new income estimate model for consumer targeting today.
The Andover, MA-based company builds its own data, which it uses in conjunction with a proprietary automated modeling process. Genalytics built its income estimate data from publicly available sources and self-reported actual incomes, which were later validated; the data was then run through the company’s genetic algorithm modeling tool.
Premium Income Estimate gauges the affluence levels of prospective customers and is particularly accurate for prospects who earn more than $75,000.
“In targeting, some pieces of information are much more important than others, and it makes common sense that if you had to sell a Mercedes-Benz, obviously the first thing you want to do is find who can afford one,” said Doug Newell, president and founder of Genalytics, of the company’s focus on income data. “Certain factors – age, income, home value, an estimate of overall wealth – a few of these variables drive many models. Other things come in to help out, but the most important thing when selling things is who has the money to buy it; that’s why this estimate is so crucial to the industry.”
Genalytics CEO Ray Kingman recounted that, in building more than 10,000 models over the course of the year, he couldn’t recall one that didn’t use income as a factor. Campaigns as diverse as luxury condos and pizza all used income data to target prospective consumers.
After building the model for its Premium Income Estimate, Genalytics created a sample and measured it against three major compile vendors. The estimates, which measure by wide bands of $25,000 and narrow bands of $10,000, showed a 39% to 67% improvement in accuracy across the three vendors.
Genalytics uses an automated analysis engine, which allows it to build localized models for direct campaigns.
“We don’t have to try to fit to a one-size-fits-all format, and we don’t have to build a national model,” said Newell. “We’re capable of building a model for every state or zip code to fine-tune to local demographics, therefore we don’t have to apply the same equation to an affluent area like Greenwich, CT, and to some part of rural Alabama, which others have to if they only build one model.”
Added Kingman, “The improvement comes from being able to get locals. One location is not diluting the scores of another location.”
The new income estimate will be used in the “Genalytics Universe” – a set of proprietary variables used by more than 100 Genalytics partners. Premium Income Estimate will also be licensed out to other companies, such as financial institutions, who are looking to add in-house data to their databases.
Genalytics is now building a financial information suite, which will include the income estimate, home value, loan value on a home, wealth ID, auto value and auto consideration sets.
“There’s a segment of the US population who, unfortunately aren’t in the position to buy much of anything, so no matter what you’re selling, income will come in to screen those people out,” asserted Kingman. “As you improve targeting accuracy, as the economics of direct marketing goes, improvements on accuracy can double ROI on mailing, so this is a big deal.”