Each year, there seems to be new buzzwords making the rounds at the Money 20/20 conference.
This year, it’s machine learning. Just like the cloud and Big Data, the term can mean different things to different people as the concept matures, but here’s one take from Ned Tobey, vice president of product management at digital banking specialist Q2.
We define machine learning as the ability to examine and analyze large amounts of digital banking data to find and recommend customer insights without a computer explicitly being told to do so, Tobey said Tuesdayduring the conference’s third day at the Venetian in Las Vegas, NV.
Tobey said analyzing trends in deposit and payment patterns allows his firm to use techniques such as collaborative filtering to assign traits to consumer activities.
It’s not just about recommending new products, it’s about not missing opportunities, Tobey said. [It’s]about fulfilling customer needs with products those customers need at the time they need it.
For credit unions, for example, Tobey said a use case could be balance alerts throughout a payment cycle. Setting up an alert that tells a member when a balance isless than$500 can be useful but not entirely contextual depending on the point in their income/spending cycle.
Instead, Tobey said, identifying upward and downward trends in spending and deposit behavior allows the algorithm to dynamically adjust and tailor an alert to an uncharacteristic balance or behavior at any given time.
Another way to look at it is to consider machine learning to be Big Data on the move, vast stores of information used across multiple channels in diverse ways with techniques and models that repeat and can revise based on what comes back.
That’s what’s happening at First Data, processor of hundreds of millions of card accounts for issuers that include thousands of credit unions through Card Services for Credit Unions, not to mention some 6 million merchants, and operator of the STAR ATM network.
Machine learning is the ability we’re now gaining to deploy intelligence on that data at scale, said Steven Petrevski, First Data’s senior vice president of security and fraud solutions. We used to have to build fraud tools for a specific channel and then do it again in a different way for another channel. Now we’re learning how to bring all this stuff together and use it across all that complexity.
Ken Myhra, director of retail payments and deposit products at BECU ($15.1B, Tukwila, WA), has an even simpler definition: Machine learning is about pattern recognition. It gives us information we can leverage to really deliver a white glove experience to our members.
Machine learning also has a close cousin that’s become a term nearly as ubiquitous and often used somewhat interchangeably: artificial intelligence. That’s also a technique that, as Fiserv senior vice president of marketing and strategy Matt Wilcox said, uses predictive analytics to learn from the data and then optimize the user experience based on that learning.
The application of artificial intelligence in use perhaps most obvious to the non-technical is the responsive language ability of such things as Erica, the chatbot announced this week at Money 20/20 by Bank of America. Using constantly changing data the big bank has on its customers, the service will dispense financial counseling advice to users of its mobile app when rolled out next year.
Another example is Abe, a chatbot that uses SMS to provide advice on how much money is left for groceries this week, as one example, and how much you’d need to save to take that trip to a friend’s wedding next month.
Conversational banking is an undeniable trend, said Keith Armstrong, founder of Abe, which he said will soon also be available on Facebook messenger and Alexa, the voice service that powers the Amazon Echo device.
Machine learning is also here to stay, it seems. Petrevski, the First Data exec, said approximately 100 people attended a Money 20/20 pre-session on the topic.
Are we going to be able to find each other at this conference five years from now? he asked. This doesn’t feel like a trend or the topic of the year. This is a disruptive force across many industries but especially finance and payments.
Also From Money 20/20
- The Hearts Of Non-Profits And the Mentality Of Hungry Entrepreneurs
- Fintech Disruptors And Regulators Can Find Common Ground