Bank of America Transforms Its Information Technology The Problem After the financial crisis of 2007-2008, Bank of American (BofA; www.bankofamerica.com) executives directed the bank to take measures to eliminate the problems that led to its near-death experience in that crisis. Many of these measures involved BofA's information technology. As the bank spends approximately USD $3 billion developing and buying information technology each year and approximately USD $9 billion on maintaining its existing IT infrastructure, these measures have taken on critical urgency. A Multipart Solution BofA developed a series of strategic goals for its IT function: (1) standardize its IT infrastructure, (2) streamline applications, (3) develop customer-focused innovations, (4) streamline the physical infrastructure, (5) develop a digital mortgage process, (6) utilize artificial intelligence in bank business processes, and (7) monitor the FinTech space for possible strategic partnerships. BofA's IT organization, called Global Technology and Operations, was directly tasked with achieving these goals. IT Infrastructure. In the past, the bank's business flexibility was limited by a huge, global IT architecture that cost billions of dollars per year to operate. For example, the IT department typically allocated separate servers in its data centers for each line of business, such as its mortgage business and its trading applications. To improve efficiency and reduce costs, the bank's first strategic goal was to create a standardized, shared IT infrastructure that all business units could readily access. To accomplish this goal, the bank implemented a software-defined infrastructure (SDI), a private cloud (discussed in Technology Guide 3) where software provides the server, storage, and networking resources to business users as needed. This process is similar to the way cloud service providers offer computing on demand to their customers. The SDI private cloud enabled the bank, which operated 64 data centers in 2008, to plan to operate only 8 data centers by 2019. The SDI private cloud enabled the bank to respond more quickly to changing business conditions and to cut costs as much as 50 percent from the bank's current data center costs. The cloud approach also enabled the bank to perform more computing tasks with less hardware. In addition, it increased the ability of the bank's network, storage, and server capacity to scale up or scale down as business conditions dictate. Essentially, the cloud gives the bank more flexibility and speed to react to changes in the bank's dynamic business environment. The SDI private cloud requires more complex security controls and compliance reporting. BofA must be able to prove to auditors that it is securely managing sensitive data such as bank account numbers and credit card information in its cloud. In fact, to improve information security and regulatory compliance, the bank is experimenting with tagging or labelling each piece of data so that it can follow the data across its global network, tracking anyone who has had access to it or has made changes to it. Applications. The BofA has retired more than 18,000 applications, many of them left over from its acquisitions of other companies. For example, the bank spent USD $100 million to consolidate five Merrill Lynch financial adviser applications into one. (BofA acquired Merrill Lynch in 2013.) Further, users at corporate clients in 140 countries now access BofA's CashPro Online portal in 11 languages. The portal replaced hundreds of applications, including liquidity management, currency conversion, wire transfers, and many others. The bank also consolidated 22 collateral management systems into 1 and 8 teller systems into 1. Simplifying its IT infrastructure and reducing the license and support costs of thousands of applications have enabled the bank to invest a larger percentage of its USD $10 billion annual IT budget on new, innovative applications. In fact, BofA has doubled its spending on new development since 2009. Innovative Customer Service. The bank's customers stated they wanted their bank to be "where they are." In response, the bank launched new versions of its customer smartphone and iPad app in 2014. The app provides three features: 1. Account information and transactional capabilities. Customers must be able to view account details and transfer funds on any device from wherever they are. Further, customers can order new debit and credit cards, view their available card credit, schedule appointments, modify scheduled bill payments, order copies of posted cheques, and perform many other functions. 2. Service. The app provides numerous service-oriented features. As one example, if bank customers are traveling internationally, then they should be able to place a travel notification on their accounts via a mobile device rather than having to call the bank to speak to a customer service representative. 3. Mobile payments and commerce. BofA is offering services such as its clearXchange person-to-person payments network jointly with JPMorgan Chase and Wells Fargo. In addition, the bank offers its BankAmeriDeals merchant-funded rewards program, which allows customers to receive coupons from retailers by clicking on offers sent directly to their online banking accounts. By the fall of 2018, the app had more than 20 million users and was growing rapidly. BofA is now managing millions of transactions made from mobile devices, an amount that is growing at a rate of 50 percent per year. Physical Infrastructure. To reduce costs, BofA analyzed its network of bank branch locations. The bank tracked every transaction by customer, location, time, and channel. For security purposes, the bank removed all details that would identify any individual. In addition to tracking customer activity, the bank examined the capacity of each branch, the costs of each location, each branch's total revenues, annual sales, and first-year revenue sales to a new customer. Based on the results of the analysis, BofA reduced the number of its branch banks by 20 percent. The Digital Mortgage. In April 2018, BofA launched a digital mortgage. The digital mortgage lets customers fill out a mortgage application through the bank's mobile app or its online banking Web site. Mortgage closings must still be completed in person. In creating its digital mortgage, the bank reduced the number of fields that customers have to enter into the mortgage application to less than half. A standard mortgage application has over 300 fields that must be completed. The bank prefills the majority of that information. The bank retrieves relevant data and documents from its own systems, saving customers from having to enter so much information and having to take photos of documents. For noncustomers, the bank works with a data aggregation firm to retrieve the necessary information from other financial institutions. The bank notes that the digital mortgage reduces the time it takes for customers to get to the mortgage closing by half. The bank's existing Home Loan Navigation software tool guides the customer through the process and provides status updates. BofA emphasizes that the digital mortgage does not replace the lending officer, but enhances the experience of the lending officer by making the busy work much easier and faster. Utilize Artificial Intelligence. The bank introduced Erica, its chatbot, in March 2018, and by July 2018, Erica had 1 million users. Erica uses two forms of artificial intelligence: natural language processing to understand speech, text, and intent, and machine learning to gain insights from customer data. Erica helps customers check balances, reminds them about bills, schedules payments, lists recent transactions, and answers bank-related questions. Customers can interact with the chatbot via voice, text, tap, or gesture. Customers' top use of Erica is to search transactions. For example, they can access Erica and say, "Show me all my Walmart transactions." Customers search for pending transactions by saying, "Did my paycheque post?" They also use Erica to obtain account balances, account numbers, or routing numbers. Further, they can pay bills through Erica. By late summer 2018, Bof A introduced insights and proactive notifications through Erica. Here, customers will be told if they have upcoming bills due, if their balances are too low to cover those bills, if a subscription to a magazine is expiring, or if their credit score has changed dramatically. The bank is handling a variety of challenges with Erica, including: protecting consumer privacy, improving the chatbot's ability to understand everyday speech, and meeting additional demands that users are sure to place on the service as they become accustomed to using it. The bank also uses machine learning in its fraud management programs. The overall goal of fraud management is to lower losses and enhance the customer experience. The basis for fraud management is a thorough understanding of customer behaviour. That is, understanding what is a normal transaction and what is not. As the bank gains additional insight into customer behaviour, it can refine its fraud management algorithms to ensure that the bank does not identify a legitimate transaction as fraudulent nor miss a truly fraudulent transaction. The bank also applies machine learning to billing disputes. Rather than an employee spending hours gathering data about a dispute, machine learning software finds and analyzes this data quickly and delivers a verdict. Machine learning also holds promise in analyzing trade data. For example, the European Commission's Markets in Financial Instruments Directive 2, which took effect on January 3, 2018, calls for additional reporting on trades. As a result, the bank is recording 2 million more financial transactions than it was prior to the directive. The analysis of those events is helping the bank see when a transaction made money or lost money. The bank is using the analysis to refine its financial offerings to its clients as well as to refine its own financial operations. BoA executives do have concerns with AI technologies. For example, there is a chance that AI models can be biased. Therefore, the bank will not allow an AI algorithm to have the final say in who is hired. FinTech. FinTech refers to the innovative use of technology in the design and delivery of financial services and products. FinTech encompasses lending, advice, investment management, and payments. Many FinTech companies utilize mobile technologies, Big Data, and analytics to customize products for various customer segments. There are many examples: Digital payment systems such as digital wallets, mobile payments, and peer-to-peer payments; Investment systems such as crowdfunding and peer-to-peer lending; and Financing systems such as micro-loans. BofA operates in a highly regulated industry and has very narrow margins. Therefore, the bank will not become a FinTech company. However, it is exploring various strategies to work with FinTech companies. Interestingly, Bank of America IT executives meet annually with 40 technology startups in Silicon Valley to learn about new products. Over the years, the bank has decided to do business with about 17 percent of these startups. BofA's work with startups balances the bank's need for scale and reliability against its need for new ideas. BofA has high expectations for these vendors. The bank requires open standards and interoperability, meaning that vendor applications must integrate seamlessly with BofA's applications. It also requires technology contracts in which costs scale down as well as up. Finally, technology vendors must share BofA's risk and regulatory rules, including, in some cases, agreeing to contracts where vendors share in the liability if their technology causes problems that lead to losses or fines for the bank. In just one example, BofA is working with Quarule (www.quarule.com), a company that applies artificial intelligence to banks' regulatory compliance tasks. The Results The IT transformation is ongoing, and the bank's financial results seem sound. In the fall of 2018, Bank of America was the United States' second-largest bank, with more than USD $2.2 trillion in assets. Further, the bank reported net income of USD $18.2 billion in 2017.