Bank of America's CTO on Technology's Future | Straighttalk

Newsletter Subscription

Keep up with new content on the site, receive exclusive content and commentary, and learn about activities within the Straight Talk community.

By Peter High, President, Metis Strategy

​This article is by Featured Blogger Peter High from his Forbes.com column. Republished with the author’s permission.

As the Chief Operations and Technology Officer of $94 billion Bank of America, Cathy Bessant has tremendous influence in the technology ecosystem that surrounds the financial services industry from fintech start-ups with which she partners, from the war for talent with the would-be employees of the company to the vendor partners that she engages to the rising influence of artificial intelligence (which she prefers to refer to as “responsible automated intelligence").

It had been three and a half years since I last caught up with Bessant, and she had mentioned that she had rethought a few things. She used to refer to Bank of America as a fintech firm. She has come to realize that it is not a fintech firm, but that fintech firms are an important part of the ecosystem that the company has built. In fact, she highlights the advantages a company with the scale that Bank of America has relative to the fintech companies who compete with aspects of what Bank of America does. She also used to note that Bank of America would never be involved in public cloud solutions, but she now foresees a day when that will be possible. She explains these changes of heart in great depth herein. She also covers the boards that she has joined in recent years and the value she has derived from them and vice versa among a variety of other topics.

Peter High: Cathy, several years ago, we spoke about your purview over both technology and operations at Bank of America. Since our last conversation, the trend toward the intersection of these two disciplines has become more profound. We see it with a number of executives who have taken on both sets of responsibilities, walking in your footsteps, if you will. We see it even at a more granular, technical level with topics like DevOps, which combines the two. Having owned these areas for such an enormous organization, what new insights do you have regarding the advantage of having them together?

Cathy Bessant: When we combined the areas seven years ago, even our own people thought we were crazy. Today, the intersections are becoming increasingly apparent. Everything we do in operations, from the most manual to the most sophisticated task, is better when it is automated. It is better, cheaper, and safer when it is digital; and it performs better for customers. For a classic operations organization, technology is not just nice to have, it is a requirement. By the same token, technology is nothing if it does not perform. The intersection of the operations team’s contributions to platform stability and to ensuring that there is not only a technological experience, but a human experience, is what today’s customer expects.

High: The rise of artificial intelligence has been profound in recent years. You prefer to refer to it as “responsible automated intelligence,” rather than artificial intelligence. Can you define the term and share the reason for your preference of that specific terminology?

Bessant: I do not like the term artificial because by definition it means fake, so it is partly a linguistic preference. More importantly though, intelligence is created by humans, even if it is run on an automated basis. We can think about artificial intelligence as algorithmic intelligence, as well, because what takes data and produces an outcome are algorithms that are written, overseen, governed, and managed by people. Automated intelligence is often better, more predictable, faster, cheaper, and has a lower error rate - as long as our algorithms work. There is no doubt AI is the term of the day, but we have been using automated intelligence for a lot longer than people think. For instance, we have been using models to create credit scores for 25 years. You can go through a lot of different elements of banking and say, "This is a more sophisticated or more tech enabled deployment of automated or artificial intelligence,” but it is not something we are not used to, which is an important point.

High: You use the word responsible when referring to automated or artificial intelligence. How do you think of the responsibility aspect?

Bessant: Models should improve decision-making; poorly constructed models can bias decision-making. Just because a model is algorithmic or automated, it does not mean it is unbiased. For example, we try to make sure that our lending activities for mortgages are not biased geographically so that everyone has an opportunity for home ownership, regardless of the ZIP code where they live or want to buy a house. If you create a model that does not have ZIP codes, you may think it is unbiased, but in reality, if you put several ZIP code proxies into that model, you can create a biased outcome. I spend a lot of time thinking about the rapid rise of artificial and automated intelligence and how we can ensure it is responsible, which is why I often use words like govern and oversight. At the end of the day, the automated or artificially intelligent vehicle, for instance, will have to make a choice: Does it hit a tree? Does it hit a deer? Or does it hit someone crossing the street? Drivers make these choices every day. To design an algorithm that would make that decision, and to think of it as being potentially irresponsible, is the great challenge that we face. It is one of the reasons that artificial intelligence will not decimate the workforce and will not make humans obsolete. While humans may not be performing the actual decision in the moment, they are certainly creating every element of the framework that gets to that decision.

High: You are a proponent of the idea that AI and data will make or break some companies. Being such a large company, Bank of America has data in many disparate places, how do you bring that data together to make better decisions?

Bessant: We are doing it better every day. Data is one of the most important natural resources of our time. How we harness it determines whether it is a rainstorm, or it is a rainstorm that produces water that produces irrigation that produces crops. I think of it in terms of how we take the big universe of data and make it usable in a way that changes product opportunities, changes lives, changes worlds, and makes things better. That takes brilliant architecture, brilliant storage, brilliant transportation of data, and well protected data; something that is increasingly, horrifically hard to do. Any outcome we want to generate is foundational to the discussion of data, and vice versa. At Bank of America, no matter how good we are at it today, it is probably our most untapped resource. We are, in many cases, in our infancy in thinking about how we can use our disparate data to drive our mission, which is to improve customers’ financial lives. First, my team tries to get crisp, streamlined, and agile data; that is the manufacturing part of data. Then, we work with the businesses so they can understand the data’s capabilities and how to access it.

High: Fintech is getting a tremendous amount of investment. Venture capital, for instance, is pouring a lot of money in this area. There is sometimes a tendency among established companies to have a “woe is me” perspective in contrast to the start-up community. However, there are tremendous advantages to being a major enterprise, including having an established customer base and the ability to draw insights and inferences from data that has been collected across decades. What is your perspective on the opportunities major enterprises have relative to startups?

Bessant: More than likely, if you had asked me this question the last time we talked, I would have said that that Bank of America is a fintech company. That is not true, we are not a fintech company and we are never going to be a fintech company. Bank of America is highly regulated and our margins are razor thin. The idea of “fail fast and fail often” does not work in a business where your gross margin in a wildly good year, might be a little bit under two percent. “Fail fast and big” is not in our lexicon. Furthermore, our shareholders are value driven. They want predictable performance. They want a dividend. When it is appropriate, they want share repurchases. A VC funding model, which can be more patient and entrepreneurial by design and is built to produce outsized returns, does not fit the way a large public company is constructed. The basic building blocks of fintech will never be ours.

That being said, big business has to be great at doing three things in this area. First, we have got to be great at working with, and buying from, fintech companies. We need to teach them how to do business at scale; how to enable technologies that work at one volume, rate, or capacity to work at others. Another part of this is enabling tech companies to wade through the labyrinth of red tape that we, appropriately, put them through. We have to be brilliant at buying from fintech companies because there are times they are better and faster, and think of something we have not. The second thing is, we must learn how to partner differently. Large companies are not always good at partnering. Apple Pay is a model where the client and customer share economics. The business model is completely different from the traditional one. By thinking about economics differently, Bank of America will be on the cutting edge of what partnerships can look like. The last of the three is that we have to be brilliant competitors. There are things that fintech companies occasionally are, or try to do that threaten us existentially. We have to use fintech to make us better in those existential moments.

There are many advantages to being big. I can, and do, invest heavily in information and data protection. We can be physically ubiquitous. In the cities where we operate there are a lot of Bank of America branches and ATMs. We can also be ubiquitous from a distribution perspective. Scale gives us diversification, which means we can we can ride out the variability in any individual idea or business. Not every idea has to work, and not every business we are in has to be wildly profitable at the exact same time.

High: You alluded to the sanctity of security and risk mitigation, obviously an important concept in any company, all the more so in an organization like Bank of America. What is your approach to risk mitigation as you innovate and push the needle?

Bessant: It is partly why we can be slower to market. Not that we like to be, but because we respect the fact that a small mistake, or a small vulnerability, across 30 million households in the United States, could be could be systemically damaging. First and foremost, we take our responsibility seriously. We are in the trust business with our clients and our customers. For me, because we have people’s money, the definition of our brand is trust.

Part of the way we think about risk mitigation is investment. Part of the way we think about it is a structured design process, which includes avid testing. Part of the way we think about it is talent. In the information security space, but also in the technology space overall, the thing that keeps me awake the most at night is talent. Whether you are designing a defense system, or executing a defense system at the point in which someone is trying to attack you, there are judgments that have to be made on a second-by-second basis. You need the best talent for those decisions. Our way of thinking about it can be summed up in four words: best tech, best talent. We are in a war for talent. I have never in my 35-year career felt it so acutely as I do today. The demand is high for great technologists who can also lead businesses and lead teams. The tech sector, financial services, and defense, we are all chasing the same people.

High: We have talked about a variety of topics that are rising in importance in technology: artificial or automated intelligence, data analytics, and security. What sorts of skills are you looking for today that are different from when you first took your post? How do you see that evolving as you think about the war for talent and building a team that supports not only the technology and your current strategy, but also those of tomorrow?

Bessant: More than anything, the difference today versus five years ago is that we have to look five years out. Whatever we thought our leadership model was, it has to be reconstructed. I have been spending quite a bit of time with my management team constructing what we call 'leader attributes of the future.' It does not matter if you are an operations person or a technologist, we are looking for naturally digitally thinking people who have a high capability to drive change, not to only execute change, but are change hungry. We need people who love change. My team has set out approximately 25 different leader attributes. These attributes will be used for everything from how we hire to how we develop our people internally. One thing that is important in a large organization, and I hate the term, is exponential thinking. Leaving behind the land of two percent, five percent, or eight percent change, and thinking about how you can take the technology or capabilities of today and stand them on their side to produce an exponential outcome.

I will pick something pedantic to explain. We spend a lot of time trying to get our customers to change from getting their bank statements on paper to getting them electronically. We set goals, move from 25 to 30 to 40 percent. However, the kind of thinking we need is: How do I ensure that all new accounts opened at Bank of America only get electronic statements? And, what does it take to get from 40 to 100 percent? It sounds easy, but it is not. You have to turn the problem on its side from how it has been thought of traditionally. This is how you make true leaps. This is what fintech companies are able to do because they are starting from nothing.

High: Another playbook item for companies that are starting from nothing is to largely have cloud-based technology. There has been a lot of debate in large enterprises, but financial services more specifically, about the wisdom of leveraging the cloud, especially public clouds. What are your thoughts on this?

Bessant: It pained me greatly, but I had to change my point of view. When we first started considering clouds, five or six years ago in earnest, and talked about turning fixed assets into variable assets and fixed capacity into variable capacity, the capability to protect oneself in a public cloud environment was not nearly what it is today, and is not what it will be five years from now. Two or three years ago, I would have said that Bank of America will never be on a public cloud. I do not believe this anymore, one day it will happen. Today though, we are fortunate that another benefit of scale is the ability to virtualize our assets on our own. We run our own cloud. We are also making forays into what I call multitenant clouds. These are clouds run by organizations like Workday, Salesforce, and Oracle. We ride these companies’ architecture and infrastructure when we need variable capacity and when we can be more efficient in a multitenant environment. These are still different from a public cloud though, where anyone who wants to pay can play. That type of environment is much tougher to secure. In the next few months, when we have the right work and when the safety and security is at a level where we are comfortable with running some initial tests, we will begin to dip our toe into the water of having some partnerships where we explore, in a limited way, using a public cloud. I cannot afford to be flat-footed at the moment when the public cloud is safe and secure, which it will be. For the time being, we will continue to be not just predominantly, but virtually, all in-house. Then, there are occasions where we will work with third parties in a multitenant environment. Eventually, slowly but surely, we and others, will drive the public cloud to a safety level that works.

High: Many people, as they manage their career, have a tendency to grow in a silo of business. For example, if they are hired as an Operations employee, they work their way up the chain in operations; the same for Marketing, Finance, or Technology. However, you have meandered across a variety of disciplines. I imagine that you draw from all of those in your current role, and that they have given you the advantage of not only depth, but also breadth. It strikes me that this is another tremendous advantage of a business at scale. For instance, if there is a high potential person who sits under a person who is not going anywhere, the junior person has the opportunity to have a set of experiences, and grow, by moving to another part of the organization. You may lose them to a different division, but not to a different company. First, is my hypothesis, correct? Second, to what extent is this part of Bank of America’s career model, providing both breadth and depth experience to talent?

Bessant: To take a broader view for a moment, historically, some of the technical or functional disciplines like Finance, Risk, Technology, and Operations, have valued that silo. Not only have organizations not been adept at creating a wide breadth of development, individuals in functional disciplines have not embraced it. There are still times when I say to someone, "Go spend three years in this business, and you will come back a better technologist, a better operating person, and you will be of more value to the firm," and it is a difficult discussion that does not get immediate buy-in. However, I could not do my job if I did not have a breadth of perspective because there is no front office and back office. Technology has completely changed the notion of business integration. You cannot say the business is technology or technology enables a business, they are one and the same.

Your hypothesis is 100 percent correct. Scale helps. The only reason I have had a 35-year career here is because there is not a discipline I have not been exposed to. Heck, I can play lawyer some days because we always have things going on from a technology perspective with our legal team. I have never had to leave the firm to get an opportunity to learn. That is a benefit of scale and something we focus on. My boss often says that part of his challenge is to keep me on a learning curve because otherwise I create crises. Like the saying goes, an idle mind can create all kinds of havoc. Truly, it is in the best interest of the firm that I have had this kind of career.

Our structure is deliberate. I am required to do a talent planning analysis twice a year. We have been doing them for 20 years. I can tell you who on my team is ready for the next development move and who is not, probably three-deep into my organization, because we are disciplined about how we look at career assignment and career rotations. I set goals for myself for both importing and exporting talent. Over the last year, I added a new head of Global Markets, Sales & Trading Operations. He came from the business; he has previously run Global Treasury Services and Prime Brokerage. Now he is running operations for both of those businesses. It is a big learning opportunity for him, and a great opportunity for us because he brings a fresh perspective as well as client and counterparty knowledge. These sorts of things do not happen by accident though; your hypothesis is spot on.

High: Since our last conversation, you have added several board affiliations. A partial list includes the boards of Zurich Insurance Group, Florida Blue, UNC Charlotte, and the University of Michigan's Ross School of Business. What advantages have you garnered from your diverse board activities and interactions with other board members?

Bessant: My board involvement makes me better at what I do, period. I also believe it makes me a more well-rounded person, which in turn makes me better at what I do. There is a real difference in people who have an outside perspective; I see this in my colleagues and peers as well. Zurich is my first public board. It is an interesting experience to look at capital and a balance sheets that are not ours, to think about customers differently, and to think about policy versus execution, which is what being on a public board requires. These things are all stimulating and they are impactful on me. I hope, as I go forward, they will be on Bank of America as well. I came back with from my first Zurich meeting with pages of notes. Some were on the company, but many were on the ideas that were stimulated from listening to the questions and thoughts of the other board members. It is a leadership lesson. It is a breath of perspective lesson. It is reflective. It also makes me better at presenting to Bank of America’s board because I know what they expect, and what they do not care about.

High: Across your board experiences, what types of questions come up related to technology? What are the appetites or curiosities about the opportunities and threats represented by a topic as grand as technology? Do you see the questions weighing more toward security issues as opposed to innovation?

Bessant: There is a pull for both. There are avid discussions at the board level around protection and security, but also a real worry across boards, and I see it in our clients’ boards as well, as to whether they are quick enough, agile enough, and are innovating enough. Perhaps the biggest risk I see across the different boards, is the belief that technology cures everything. If we go back to the responsible AI discussion, it is not enough to have technology, it has to be great. I do worry about that, the belief that more is better in the technology space. Sometimes the discussion is, let's flip the switch now. Before I came into this position, and this was having run several businesses at the bank, and for years having what is the hardest job in the company, chief marketing officer, having done all of the wide variety of things I did and all of my knowledge, I still marvel at how I thought you could flip a switch and make a change in technology. And how much I thought you could guarantee operational perfection. What have I learned? There is no switch that can get flipped that is instant change. Also, there is no such thing as foolproof. It always does, in fact, come down to the guy in the data center who trips over the cord and pulls it out of the wall. The complexity of what we try to get right every day is why I have so much respect for my peers. The nonstop need to be thinking about resiliency, about Plan B, and about talent, are all more complicated than I thought before taking this position seven years ago.

Originally published on Forbes.com