Emerging marketing technologies: Part I

How technology transforms marketing

'Supervised machine learning' requires patience.

February 12, 2019

Driven by technology, marketing has been transformed within the past decade. Mobile technology and connectivity drive commerce—and consumer demands—around the clock.

Computing power is virtually limitless. Data and analytics shape a retail journey that has moved from meeting consumer needs to anticipating them.

In a three-part series, we examine how emerging technology is reshaping today’s credit union marketing function.

First up: Artificial intelligence (AI).

Increased computational power has provided organizations and individuals with the ability to use artificial intelligence to extract patterns and insights from huge amounts of data that previously weren’t accessible.

Companies are ready to use AI to create a competitive advantage. According to a survey from MemSQL, 61% of companies, regardless of size, cited machine learning and AI as their companies’ most significant data initiatives.

Faraday employs an AI platform to help credit unions “convert, engage, and retain” members. That platform compares a credit union’s membership data with what it calls an “identity graph, a database of close to 260 million Americans, that includes demographic, psychographic, lifestyle, financial, and property attributes.”

“We’re enriching our understanding of who the credit union’s members are, sometimes including people who expressed an interest in membership but didn’t necessarily become members,” says Robbie Adler, Faraday co-founder/chief strategy officer. “We can use that information to uncover patterns.

“We can ask, ‘Who are the people most likely to become members?’ or ‘Who are the people who not only have deposit account, but also have an auto loan?’ The idea behind machine learning models is trying to get beyond a rules-based or segmentation-based approach to marketing.”

After identifying these “audiences,” as they are called, the Faraday platform makes marketing and communication platform recommendations, such as social media, direct mail, or e-mail. “We might say, ‘Hey, here's a great group of people for you to market to on Facebook.’ Or you can download a direct mail lists if it's an existing membership database. We may push those recommendations into an email tool or a marketing automation system like HubSpot.”

A learning process

Roger Rassman, vice president of marketing at Community First Credit Union in Jacksonville, Fla., was “guardedly optimistic” when the $1.6 billion asset credit union implemented Faraday’s platform about a year ago.

“Implementing artificial intelligence takes time,” he says. “The platform has to learn.”

But Rassman is patient with Community First’s AI efforts. The credit union uses “supervised machine learning” on several products.

“Supervised machine learning is when we tell the machine what we want the outcome to be,” Rassman says. “Unsupervised is when the machine goes through all of your data and tells you what the outcomes are.”

The credit union will, for example, request a list of likely auto loan candidates from Faraday, to which it will market through digital channels. The credit union tracks those results and sends the information back to Faraday, which appends the various attributes and behaviors in its database.

“That’s how the machine learns,” Rassman says.