Most companies realize their IT systems are overly complex, having been added to multiple times over the years without much effective pruning. The most common response is to simplify systems using new technologies like the public cloud and SaaS platforms, which help to gain speed and lower costs. But those are temporary fixes. Complexity inexorably creeps back over time, due to the fundamental nature of people and organizations.
In the long run, managing IT complexity is a function of changing behavior. After-the-fact technical fixes need to be complemented with an approach that seeks to curb complexity before it is created. “[The biggest risk] is that business reverts back into the ‘I want this, you [IT] go and do it’ mode, the old pecking order,” says Jeroen Tas, EVP and CIO, Philips. “It’s not that people don’t buy in on the logic, I think they do. They revert back to their old behaviors, because they are deeply ingrained. So the biggest obstacle is mindset.”
Based on interviews with 35 IT and business executives on the challenge of addressing complexity, and multiple in-depth case studies, we have identified three levers that help drive a behavioral approach to blocking IT complexity (see Figure 1). The methodical framework we used in the interview process derives from game theory, which focuses on the sub-optimal outcomes that result from asymmetrical information and power in decentralized groups, which applies aptly to the business-IT relationship in many enterprises.
Figure 1. Ways to deal with IT platform complexity – a framework
1. Create transparency of the consequences of business decisions on the complexity of digital platforms.
Executives need to create transparency by removing information asymmetries between those creating platform complexity and those who have to deal with it. Most best practices support creating a common understanding by bringing together cross-functional teams, including IT, product management, customer service, and others. For example, IBM and USAA rotate executives in and out of IT; Pat Toole used to be the CIO and then became the leader of a business unit. Those leaders now know what complexity does to IBM and act accordingly in their new role.
Much of the research on IT platform complexity has tried to come up with extensive metrics for architectural complexity and its drivers. These include the size, diversity, integration and rate of change of IT artifacts on various architectural levels like IT infrastructure, applications, and information. Decision-makers need to understand the implication of choices based on those metrics. For example, Akzo Nobel is moving from over 180 different ERP systems to only six. And Philips is replacing its 10,000 systems with a few hundred that form its Philips Integrated Landscape, supporting all core business processes.
2. Create a compelling motivation for avoiding platform complexity.
Companies can set incentives (or penalties) that address platform complexity. A focus on individual business cases and short-term departmental or business unit performance encourages local optimization that often increases enterprise-wide platform complexity. In contrast, leaders in stamping out complexity encourage enterprise thinking—by adapting their bonus and performance evaluation structures, by encouraging executives to model behavior supporting enterprise-optimized decisions, and by setting policies that favor the whole enterprise over individual business units.
For example, at ING Direct Spain and USAA, corporate-wide incentives count at least as much as local unit incentives, encouraging employees to think about enterprise-wide impacts (including on IT) rather than just their local impacts.
3. Redress the power imbalance between those creating and those having to deal with complexity.
Even when companies employ cross-functional teams, if managers insist on using non-standard systems, little will change in underlying platform complexity. Hence, model companies create and empower roles charged with protecting platforms from complexity (in both IT and the business units). Besides being part of cross-functional teams at the operational level, these positions respectively are also included in executive-level decision-making forums. And policies such as “reversing the burden of proof” force managers to have to argue their position a priori if their proposals will contribute to increased complexity—effectively granting more power to those having to deal with complexity after the fact.
The CEO of Bayer Material Science empowered his CIO to reduce the variety of ERP systems around the globe, as part of their “Program One,” by “reversing the burden of proof”: Instead of having IT argue with every country manager on whether they could adapt to the global company standard, country managers had to adopt the new global standard by default, unless they could prove that deviating from the standard had more benefits for the company as a whole than not doing so.
Companies that go beyond technical solutions to address complexity after the fact—by pulling on these three organizational and behavioral levers —will find themselves better positioned to compete in the digital age. They will protect and enhance their digital investment, and prevent it from becoming their legacy headache of the future.
John Boochever is a senior partner in Oliver Wyman’s Digital group and founder of the firm’s Strategic IT & Operations practice. He has thirty years of experience advising senior executives across the range of organizational, operational and systems complexity issues globally.
Dr. Martin Mocker is a professor of business administration and information systems at ESB Business School, Reutlingen University (Germany) and a principal scientist at MIT Sloan’s Center for Information Systems Research (CISR). He has researched IT and business complexity for more than ten years and is co-author of “Revisiting Complexity in the Digital Age” published in MIT Sloan Management Review in 2014.
The companies mentioned in this article derive from Dr. Mocker’s research at CISR, and not Oliver Wyman client work.