I recently had a casual conversation with a local banker about what I was working on, and I am sharing parts of the conversation. This is not a story about AI products or specific applications, there are plenty of those available, including one on our website, so you will not see any charts, graphs or links. This is a recognition that we, as consultants and technology providers, do not do a good job of cutting through the technical jargon and specific applications to explain how a local bank in Minneapolis, Omaha, or Louisville could be impacted by it.
Note: Please suspend judgement of the buzz factor and think about how they will change the way we conduct business, acquire customers, and develop products. It is worth noting that this is only one of a thousand scenario’s that could transpire and was simply for discussion.
The conversation proceeded as follows:
Banker: So, what are you working on now?
Me: I am working on how the use new technology like artificial intelligence, robotic process automation and blockchain will change the way we bank and how banks can respond. I think these will have wide ranging impacts on all banks regardless of size or location and embracing them will be vital to survival.
Banker: Are you saying they will put us out of business?
Me: No, I think banks will put themselves out of business, because they cannot afford to invest or will delay investing in the technology until it is too late (frog on the stove scenario). Once, a few banks begin to use the technology, they will differentiate themselves from the others resulting in better products, improved recommendations, differentiated underwriting, and faster, more efficient customer acquisition and processing, which will result in one of two scenario’s: 1) the best of those left behind will be acquired and 2) the remainder will slowly see their assets drained away to the point that it is no longer profitable to remain in business.
Banker: Can’t we just buy the technology to be competitive when it makes sense? That has worked for us in the past.
Me: In recent years, relying on others for technology, like a website or RDC has been an effective strategy to keep up with the competition, but I would classify those as “linear” advancements. I see linear advancements as new technology that quickly becomes relatively homogenous in nature. For example, it is possible to build better website functionality, but upon release the vendor that builds your websites will introduce the functionality within a few releases, bringing everyone back to the mean. True differentiation in linear technologies is difficult and provides marginal short-term benefit.
What we are going to experience with technologies like RPA, AI, and blockchain is what I would view as “exponential” advancement. My view is that as these are developed and implemented, the banks that use them first will quickly begin to outpace the others thus producing a chasm that will be extremely difficult to overcome, unlike our earlier experiences with technological change. Everyday, the initial gains, although small at the start, compound producing ever greater insights, capabilities and differentiation. Do not misunderstand me, there will be vendors that provide the tools or full solutions, but I do not think this will be anywhere near as easy to outsource as web development. Even if they are robust and easy to use, the early adopters of the technology will still far outpace laggards. Depending on the time delay, it may be possible to catch up, but at what cost? In banking terms, think compound interest.
Banker: Give me an example of how this would work.
Me: Let me walk through a simple example of how I envision this could transpire. To focus this example, I am going to purposely ignore the blockchain.
First, we introduce RPA in the operational units for routine entry of information from structured documents or other platforms. The technology increases scalability and input quality while lowering costs and it works 24x7. The people that used to do this work are now free to focus on higher value activities. Next, we layer in artificial intelligence for unstructured information. The software is “trained” initially by the employee, in the routine course of business, by “observing” their activities and it begins to build the new capability. As the quality of robot simulating the task increases to acceptable thresholds, the work is slowly transitioned to the robot, leaving behind just the highly complex work and substantially fewer staff performing it.
Next, we leverage the gains to focus on AI in marketing and the hyper-segmentation of customers and tailored campaign messaging. Today’s digital consumers have been “trained” by the best technology companies and expect us to know them, understand them, and provide personal recommendations and insights through their preferred channels, when they want it at the lowest possible cost. To compound this change, we have a post millennial digitally native generation, Z, that is coming of age and has even higher expectations, is technology savvy, and does not display high brand loyalty.
The response rates from the marketing programs begin to increase, taking deposits and loans away from banks that are waiting to see how it progresses. (For insight, review the use of AI in social media marketing, where is has produced a 50% or more improvement in response rates over employee managed programs). The new program, not only gathered more assets and increased loan acquisition, but also increased feedback loops, which enables us to further improve the models with each successive generation. This cycle continues through advisory services, product development, fraud, underwriting, customer service, and the remainder of the bank functions. With each implementation and generations of improvement, we further separate from the other banks, to a tipping point where laggards bank can no longer catch up and are faced with one of the two previous scenario’s.
