These causes lead to three common complexity patterns: data landscape complexity; poor business-IT governance and IT process complexity; and application and IT-infrastructure landscape complexity. (See Exhibit 2.) Not surprisingly, the extent and nature of each varies by both company and industry. Financial-services companies have greater experience with managing these types of complexity and thus handle them better. For example, most banks and insurance companies, which started to renovate their IT landscapes several years ago, score 13% better overall than other companies.
Data Landscape Complexity. Companies struggle more with data landscape complexity and data landscape management than any other issues. Approximately half of the companies in our sample scored lowest in this area. Data simplification requires uniform definitions and common standards for enterprise-wide data objects (pieces of information that are relevant to the entire company), but we found that such definitions and standards are often only partly specified; this was the case for 45% of companies. In addition, only about 10% of companies have established a “single source of truth” for their data, that is, unique storage spaces, or databases, for enterprise-wide data objects. The lack of standards and multiple databases can lead to serious problems: inconsistent data, such as data from multichannel customer engagements, hobbles the ability of data analytics programs to analyze customer behaviors and deliver digital services.
Many companies also lag in establishing clear data-management and governance policies and practices, such as defining a clear target-data landscape and establishing a central and overarching data-management role within the organization. These are important foundations for data-simplification programs. Good practices include implementing state-of-the-art data quality and security management processes and prioritizing the performance of such processes on both the business and IT agendas. Clear data-management and governance policies are not IT-only tasks; the entire business organization needs to own them.
Business-IT Governance and IT Process Complexity. The complexity of business-IT governance and IT processes was rated high or medium (at least 50 on our 100-point scale) by 60% of the companies that participated in the assessment. This pattern is common in companies with low business-IT collaboration and appears mainly in organizations where IT is seen as a service provider—a “butler of the business”—rather than as a strategic partner. The business-IT interface usually does not follow an established process (much less best practice); instead it takes the form of a complex set of ad hoc and decentralized interactions. This typically results in redundancies and workaround solutions. New business functionality is not weighed against the incremental complexity it inevitably adds to the application landscape and IT processes, such as numerous interfaces and increased testing efforts, which reduce speed and agility for the company as a whole.
Complicating matters further, if IT development processes are perceived as slow and bureaucratic when it comes to implementing new business requirements, impatient businesses try to either shortcut those processes by reaching out directly to developers to get things done more quickly or start building their own solutions, a phenomenon often referred to as “shadow IT.” This multiplies the complexity increase and compounds the problem, especially when solutions that do not comply with the overall IT architecture guidelines are employed, thus undermining digital speed or creating a security hazard—or both.
In addition, many companies do not make adequate use of agile working methods and processes. Traditional software-development methods, such as the waterfall model, tend to dominate. Although, in our experience, three-quarters of companies use some agile methods, the vast majority have only just started agile pilots: only 5% or fewer of the development teams that we encounter work in an agile mode. Introducing agile development methods and ensuring that a dedicated business owner works with the development team can significantly speed up IT delivery. (See The Power of People in Digital Banking Transformation, BCG Focus, November 2015.) Some companies might also consider applying digital to simplify IT processes—for example, by deploying code to cloud environments, fully automating delivery processes, or using robotic-process automation to monitor data center operations.
We have seen companies achieve high productivity gains through joint business-IT efforts to improve IT processes, speeding up incident management and application development, for instance, by 25% and 30%, respectively. New methods of collaboration between development and operations (such as DevOps) can speed time to market by 20%. Improving business-IT governance and thereby strengthening the role of IT and its relationship with the business—for example, by ensuring that IT has a strong role in new product-development processes and in prioritizing the IT project portfolio—is yet another way to eliminate this cause of complexity.
Application and IT-Infrastructure Landscape Complexity. The third major problem area is the application and IT infrastructure landscape, where complexity constrains companies’ ability to implement digital innovations quickly and efficiently. Nearly two-thirds of the companies in our sample evidenced weaknesses in such capabilities as developing application program interfaces and microservices for connecting new services or fostering methods such as continuous delivery (the automation of the entire application development and deployment process) or deployment.
Specific IT infrastructure capabilities required for digital services and innovations include the adoption of continuous delivery, the availability of real-time-capable infrastructure components, and the use of adequate cashing and buffering systems, such as buffer databases, to reduce latency time and processed-data volume. These are the capabilities that enable digital-native companies to use continuous deployment and other methods to release new services or software more than 100 times a day. Older, more established companies can also develop and use such resources to reduce the time for provisioning new software from weeks to days or even minutes.
One European bank furthered its digitization efforts by overhauling its application infrastructure. It designed an IT application architecture around three main domains that corresponded to three key functions: client relationships, client services, and transaction processing. The bank was able to develop applications for each domain much faster by using the most appropriate methodology in each case. It reduced the cost of developing applications by one-third while also shortening its development schedule.