I vivdly remember the lively discussions between my Mom and Dad about buying our first color tv.  Mom, who was home all day and had kids to entertain, saw the immediate benefit of color cartoons.  Dad, who was away most of the time, saw no problem with watching football and baseball games in black and white on the weekends.  Ultimately, we got our color tv, but it took a lot of back and forth.

This is where our industry is relative to the big technology changes taking place.  I'm talking about things like Payment Hubs, which I wrote about just recently, cloud-based systems, blockchain, and artifical intelligence (AI) to name a few.  It seems whenever I write about infrastructure, invariably someone will ask me something like, "How are organizations initiating these strategies?". "How much of this is hype and what's really going on."?  That's why an article in Computerworld caught my eye this week.  It rightly called out AI as one of the most important new technologies of this decade, but cautions institutions that the use case(s) for it need to be specific and clearly understood before development begins.  And that's where the problem lies.  Let's look at the best use cases defined in the article and break them down a bit:

1.  "Chat bots and virtual assistants used by retail banks to answer mundane customer questions."

This is the low-hanging fruit and one that, according to a recent survey I did for a leading call center, led the pack on both current and future investments from financial institutions I spoke to.  Not the most sophisticated implementation of AI, but it serves a real need to help institutions appear more digitally engaged with their customers.  To me, this is one evolutionary cycle away from telephone banking, but it's a start.  The ROI on this is improved productivity/reduced headcount, but that's speculative at best since it's based on estimates of shifts in consumer behavior. 

2.   "Robotic process automation or rules-based scripts that can pull data from multiple systems to generate forms or invoices."

Next level difficulty which requires an integration layer(s) to make this happen, but higher value as well.  They key is to identify the forms or invoices or other documentation that needs to be pulled together in a dynamic manner that has the most value or is in high demand.  This service could potentially generate incremental revenue, but likely more of an efficiency play in the near-term.

3.  "Natural-language processing and generation, enabling systems to read text in contracts to pick out key clauses (and determine the implications of that text) as well as enabling the system to write in plain language."

This kind of use case requires a great deal of testing and presents real risk to any organization since you'd be relying on an algorithm to interpret terminology.   One application cited in the article is the use of AI to read through expense reports, find errors or gaps and generate communication back to the employee.  Once again, the ROI on this improved productivity/reduced headcount, something not insignificant in very large, distributed organizations.

4.  "And cognitive analytics, which can find customer trends to determine which products they're more likely to purchase."

The golden ticket for most new technologies in financial services is generating more revenue through cross-selling.  AI holds the promise of providing a true map of the customer journey, something financial institutions have been working on for decades.  Of course, this is the hardest leg of the implementation and right now, something only a bank with significant resources can even hope to accomplish.

What's the takeaway from all this?  That infrastructure is hard?  No, we know that already.  What's really hard is defining an implementation path that makes good business sense and then investing in that vision. This is not different from any other leap we've made - from the first debit cards at the POS, online banking, neural networks, and mobile payments.  What is different is the fact that so many stakeholders view these technologies as competitive differentiators and so lock their learnings up in closed rooms.  The application of these technologies, in the right hands, can offer a market differentiator, but let's review once again most of the use cases in this article.  They're primarily efficiency plays and while these offer incremental profitability improvement and perhaps a short term "new feature" talking point, raising the tide of information about the real experience of integrating, testing, and implementing these technologies will serve to improve and strengthen the market as a whole. It's what the industry really needs to know. 




E-mail me when people leave their comments –

You need to be a member of KEENonPayments to add comments!

Join KEENonPayments