AI Innovation in Banking Is on the Threshold of Significant Advances

Fintechs take note, traditional financial institutions are getting their innovation act together, particularly in the use of artificial intelligence. By incubating solutions internally and then tapping outside expertise as needed, they can approach a Google-like 'AI loop' of more data insights producing more growth producing more data.

Fintech and enterprise artificial intelligence entrepreneurs sometimes can’t believe how slow banks are to innovate. They assume heritage brands, though enjoying the benefits of name recognition and customer loyalty, suffer from old infrastructure that leads to inexorable decline.

But that brand promise of longevity and durability can also accommodate innovation, particularly making financial services work faster and more efficiently. The process starts with internal experimentation. External partners can then upshift the transformation. Critics who view banking brands that predate the cloud as dated may be surprised in the near future.

The Quiet Revolution in Banking

Startup executives and challenger banks should be aware that financial institutions today are quietly researching the innovation they need using in-house data scientists and open-source software. Since they don’t know what they really need until they try it themselves, they’re incubating solutions. After all, why should they transfer their domain expertise to third-party tech companies and pay for the privilege? As a result, allegedly slow-to-change banks are going to devise notable successes.

As that evolution unfolds, the advantages of learning internally may eventually decline for some institutions. In-house data scientists must put out fires during the normal course of business. Open-source software only goes so far and imposes limits on proprietary advances. Banks that hire more scientists and technicians on the promise of discovering the Holy Grail of fintech will face reckonings. On a granular level, it’s not their expertise.

Bank executives need to be prepared to transition from a homebrew solution to partnering with a more agile player.

Achieving the right balance requires a conceptual shift. Financial institutions in reality are not facing a black and white choice between buy versus build, or between subscribing to software-as-a-service versus cultivating the internal talent to architect a new solution. They’re facing the task of establishing what Ash Fontana at Zetta Venture Partners calls a “loop moat.”

If the Bank Is a Castle, Where’s the Moat?

Executives familiar with Hamilton Helmer’s “Seven Powers,” will recall that he describes the seven ways businesses can build competitive edges that keep other companies out of their markets, allowing them to dominate. This barrier to competition is typically called a moat, just like the one around a castle.

While keeping intruders out, a moat also protects the people inside the castle, freeing them to exchange ideas and energy, creating a loop or virtuous cycle that keeps the castle — or bank, to end the analogy — thriving and relevant. In the world of artificial intelligence, a loop refers to a system where data generates insights that in turn produce more data and more insights.

Protecting one’s business with a moat ensures survival. Revving up a dynamic loop that interprets a sustainable harvest of data and sows the seeds for more information later ensures innovation. Google has predicated its entire business model on a such a loop. AI makes the loop happen.

AI Is Not a People Replacement

Some bank executives will wonder whether AI is supposed to replace people. It is not. As Fontana explains, the point of AI is to create a good user interface or other super easy way for humans to teach machines what they are doing and why.

The machines then draw lessons from those workflows and provide humans with more options to do them better. As legendary chief information officer Paul Strassmann pointed out, the ability of computers to store and use knowledge capital is their real, underappreciated value.

If machines can accomplish those same tasks and not mess them up, they have likely highlighted other duties that only humans can accomplish well. In this scenario, humans grow into the latest domain expertise where they can excel while the machines take on the jobs that were holding the humans back. Domain expertise expands. Greater worker satisfaction is often the result.

Read More:

Webinar
REGISTER FOR THIS FREE WEBINAR
How to Survive and Thrive in 2021 and Beyond
Join Shane Evans, Chief Revenue Officer at MX, and Corey LeBlanc, Chief Digital and Innovation Officer at Origin Bank, for a 45-minute knowledge-packed session.
Thursday, December 17 @ 2PM EST

Financial Institutions Are Looking for the ‘Loop’

In their quest for technology acceleration, banks and credit unions need to understand the dangers of hunting for that moat-making loop internally for too long. An institution’s data scientists using open-source tech have domain expertise, but they might not have the scalable technology or innovative teams necessary to create game-changing internal solutions that deliver over the long term.

Banks need to know when they’ve gotten the most out of their internal learning and building process, and when it’s time to transfer the right parts to an outside expert. This involves:

  • Educating themselves on the AI options available to them and auditing their current data science operations.
  • Assessing their research budgets.
  • Reassessing their goals to determine if they are hitting targets they set for their data scientists in the past.
  • Studying the competition in both the legacy and challenger spaces.

Another Option: Acquire the Innovation

The takeaway is that banks are not slow to innovate. They are cautiously deciding how to innovate smarter. However, financial institutions are also on the cusp of making crucial decisions, and they are under pressure to reach some conclusions.

The Economist Intelligence Unit found that only 15% of banks and insurance companies utilized AI, but almost 90% planned on increasing AI-related investments in fintech over the next five years. Asked why, they said AI was unlocking new growth and shrinking costs. Half said AI would drive new products and open up new markets.

Some of those investments will go to existing and new data scientists. Much will go to up-skilling workers who become significantly more productive through fintech. But the lion’s share will probably look more like French banking giant Société Générale’s acquisition of challenger bank Shine.

Shine caters to small businesses and the self-employed, functioning not only like a bank but generating invoices, receipts and tax documents while also keeping track of paperwork and administrative tasks. Shine’s users can obviously benefit from access to a major bank. Société Générale now has somewhere to steer customers who work independently or in the gig economy. That’s innovation overnight.

Many traditional financial institutions, it’s true, rely on their primary tech provider for innovation and new products. They can still do so. They own their data, after all. They can implement new strategies using loops that analyze their data. The 90% of banks studying AI aren’t ditching their preexisting systems. They’re looking for options.

Moving forward, the challenge in the enterprise AI space is to put the incoming capital and bank subscriptions to work in a way that’s consistent with their goals and interests. We can do that by finding the sweet spot for fintech, the right mix of local knowledge and outside spark that can give traditional institutions competitive advantages.

This article was originally published on . All content © 2020 by The Financial Brand and may not be reproduced by any means without permission.