- Industry: Financial Services
Originally, capacity constraints were based on how fast and accurate an accountant could complete manual math. With VisiCalc crunching the numbers near instantaneously, CPAs shifted their focus to growing capabilities and expanding professional offerings (like multiple-scenario models, “what if” assumptions, and projected growth models). Changes that originally took weeks to address now took a matter of minutes. Costs reduced, revenues grew, and the accounting industry didn’t implode. Sounds like an easy process, right? Not really. To successfully introduce VisiCalc to the industry, firms had to redesign jobs and realign expectations.
Redesigning the job
At a high level, VisiCalc necessitated separating the accounting and financial industry into functions: data entry, number crunching, and analysis. Companies that adopted this new technology went from having humans handling every aspect of accounting (picture green visors and adding machines) to sharing the work with a machine. The humans entered data, VisiCalc crunched the numbers, and humans analyzed, interpreted, and presented the results. As technology advanced, VisiCalc (later Excel) aided in the analysis and interpretation with probabilistic models and predictive analytics, but the human ultimately used cognitive ability and discretion to make real decisions. Leveraging this technology, however, required people to consider how it fit in their operational strategy.
The financial services industry is once again on the brink of a technology shift, this time in the form of automations (robotic process automation, chatbots, and the like). And just like the accounting industry in the 80’s, banks must consider job (re)design to maximize the benefits.
Completing a jobs interview
A jobs interview (different than a job (singular) interview), involves first understanding, blending, and matching skills with delivery expectations. It involves asking the question, “what are we trying to accomplish and what steps and skills are required to get there?” In general, banks haven't really done this in a long time. They haven’t needed to. The skills required of a banker haven’t changed much in the last several decades. The customer expectations, however, have changed. Between an oversaturated banking environment and a diverse competitive landscape (with alt-finance growing in popularity), a gap now exists between a bank’s skills and a client’s expectations. Redesigning what a banker could do and what a computer could do has come to the forefront of digital strategy in this industry.
From an automation lens, bank functions can loosely be categorized by whether they are rules-based or discretion-based. That is, are decisions made based on rules (credit policy, regulatory requirements, etc.) or are they made using a banker’s discretion (sales pursuits, analysis interpretation, “fringe” lending decisions). As we have found in our own automation efforts, it is more effective to think about automating activities and functions, not entire jobs or teams.
Consider the following breakdown of the traditional front and back office functions predominant in today’s banks and how automation can support, augment, and redefine banking (see figure below).
It is unreasonable to expect that automation can replace jobs one-for-one without any meaningful consideration (not to mention that automating ineffective, lethargic, or outdated processes yields minimal value at best). Job, and process, redesign is a more thoughtful and elegant way to automate. Evaluated and implemented this way, automation is a revolution, not a dissolution, of jobs.
Don't replace jobs—augment them
The generally accepted approach to evaluating tech performance is asking how good is the tech? ”Good” can mean any number of things, from accurate to fast. We propose another method of evaluating automation tech, in this case related to how machines can empower humans. Instead of, “how can a robot do this job?”, think, “how can a robot help me do this job?”
When making the case for augmenting jobs, consider the application of artificial intelligence (AI) in dermatology and oncology. A recent university study utilized deep learning techniques to train an artificial intelligence solution how to identify skin diseases. Data scientists loaded nearly 130,000 images representing more than 2,000 skin diseases and cancers into the learning model. With this information, the learning model correctly identified disease and cancer approximately 92.5% of the time. Compare this to the dermatologists and oncologists who correctly diagnosed at a rate of approximately 96.6%. While AI may be faster or more consistent, accuracy (in this case) suffered. Realistically, a case can be made for both the human and AI to do the job. But what about both?
As it turns out, the people in this example were better at identifying false-positives while AI was better at identifying the “hard to tell” cases. Combining both methods (AI and human diagnosis), accurate diagnosis jumped to more than 99%. While banking and medicine are inherently different industries, this example supports the case for embracing technology with a cooperative approach.
Creating a win-win-win
In recent years, win-win-win propositions have become increasingly popular in group dynamics. In the case of automation, the three “wins” refer to companies, their employees, and their customers. By leveraging automation, companies can realize financial benefits, employees can focus on more meaningful and impactful work, and can in turn deliver an enhanced, consistent, and positive customer experience. Bottom line: Focus on strengths and repeatable advantages and elevate your experience.
Embrace automation and let humans innovate. As technology’s presence in society increases, human interactions are reduced by choice, not necessity. While technology, automation, and artificial intelligence are incredibly beneficial resources, they are not one-for-one substitutes for good old fashioned human interaction (at least not yet). Make the most of both humans and technology by leveraging the strengths of each, aligning functions to each group as appropriate. And finally, remember that RPA software is the tool. The design is the solution and proper planning leads to effective execution which leads to long-term, repeatable benefits.