Key Takeaways from Money 20/20

Like any conference, Money 20/20 had ample educational opportunities through speaker sessions, vendor demonstrations, and open discussions with vendors on the exhibit hall floor. Through vendor meetings and attending sessions, I did learn a number of new things, some of which are shared below.

Key Takeaways from Money 2020:

• With a keen interest in artificial intelligence (AI) and its potential impacts on the financial services business, I spent a good deal of time meeting with a number of vendors to discuss these impacts. The biggest lesson I learned about implementing any type of AI solution is that financial institutions need to have access to data repositories and should have a clear understanding of the use case for the system. Having a strong understanding of what the financial institution wants the data stream and the AI analytics to produce, leads to successful implementations. Organizations that install the AI capability without a clear understanding of what they want, tend to have more difficulties with AI implementation. The other lesson I learned about AI, is that no single vendor covers the entire spectrum of AI capabilities within an organization. Vendors tend to specialize in one or two uses of AI.
  
  

• Sessions were split into a variety of different themes such as Alternative Lending or Human Factors. Within these themes were presentations geared around topics important to the theme itself. On the first day of the conference, I attended a workshop within the Human Factors theme that focused heavily on diversity and inclusion within organizations. One of the key takeaways for me was that empirical studies suggest that more diverse teams and organizations tend to have six times the innovation capability versus more homogeneous groups.
  
  

• Another Human Factors-themed session, given by a behaviorist, focused on how organizations should design their human interactions to improve financial outcomes.

A couple of interesting issues were brought to light within this session. For instance, the government spends nearly a billion dollars on financial education for impoverished and economically disadvantaged people with less than 1 percent of this population actually changing behavior as a result of this education. Organizations need to use behavioral design that clues in on psychological factors to help find ways to remove obstacles to how customers or members do things.

The speaker provided an example of a credit union that noticed a high percentage of members that simply brought their paychecks to the credit union and cashed the entire amount without putting any money into a savings account. Observations and process mapping indicated that cashing the checks was the easiest and fastest transaction for members to perform. Most other transactions required members to fill out some kind of form such as a deposit slip to make the transaction work. To change the payroll cashing behavior, the credit union required all transactions to use a deposit slip. After implementing this process, 15 percent to 20 percent of members started saving a portion of their paychecks increasing share balances across time.

• My last key takeaway from the conference is in regards to underwriting. While automated underwriting tools have been around for a number of years, some financial institutions are turning to AI and machine learning (ML) tools to speed this process further. The use of AI would offer the ability to dive deeper into available data sets providing what many institutions believe is a stronger decision-making tool. The speakers in this presentation indicated that data from credit bureaus was still the primary source of the data they use, but that the AI tools are able to go deeper into the vast amounts of data credit bureaus have on consumers and perform these tasks in a matter of seconds.

  One of the speakers in this session indicated that their credit card underwriting takes an average of seven seconds to perform including deeper dives into the credit bureau information. In the past, it would take a human underwriter 20 to 30 minutes, or more, to dig into these records and make a decision. AI as an underwriting tool for consumer loans and credit cards has a strong use case.

While there were numerous other sessions that I attended at the conference that also provided useful information, I found the above examples to be very interesting and most relatable to the banks and credit unions I work with. I feel strongly that financial institutions of any size would benefit from attending this particular conference.