Better Management of Complexity in Business

This Harvard Business Review article, authored by Gokce Sargut and Rita McGrath, outlines the concept of complex systems and how complex systems should be managed differently from complicated systems. The complexity of business environments is making business operations more difficult. Complexity has increased as more and more systems have become interconnected and interdependent mostly due to the information technology revolution.

In a separate publication of a study that was commissioned by COSO, entitled Report on the Current State of Enterprise Risk Oversight: 3rd Edition, the NC State University’s ERM Initiative found that 88% of respondents believed that the volume and complexity of risks have grown moderately, extensively, or by a great deal in the last five years. In light of these findings, it is important that businesses think about how to better manage complexity in the business environments.

Complex systems bring the following challenges to businesses:

1. They are far more difficult to manage than complicated systems.
2. It is difficult to predict the outcomes of these systems
3. They are more difficult to understand especially because of managers’ limited cognitive abilities.
4. It is difficult to “place bets” because there is no predictable correlation between the past behaviors of the system to its future behavior.

The authors highlight that there is need to distinguish between a complicated system and a complex system. One should manage complex systems differently from how one would manage complicated systems. Failure to distinguish between the two could result in expensive losses. A complicated system has the following characteristics:

1. It has many moving parts.
2. The parts operate in patterned ways.
3. One can predict the outcomes of a complicated system.  Flying a commercial airplane is an example of a complicated system.

A complex system is characterized by:

• Features that operate in patterned ways.
• Interactions of the features or parts that are continually changing.
• A complex system has the following 3 properties:
  1. Multiplicity:  the number of potentially interacting parts.
  2. Interdependence:  the connectedness of the parts.
  3. Diversity:  the degree of heterogeneity.

Air traffic control is an example of a complex system. It is difficult to predict the outcomes of a complex system. Therefore, managers of complex systems have to be prepared to handle the two challenges that complex systems pose:

• Unintended consequences, and
• Difficulty of making sense of a situation.

Unintended Consequences, which can be either positive or negative, can result due to:

The second challenge of complex systems is the challenge of making sense of a situation. This results because of the following:

1. Unintended interactions between parts of a complex system. The authors gave an example of how Nintendo’s decisions to make a simpler game console for the Wii succeeded in attracting new customers. Unfortunately, this short term success limited the number of new release titles that third party developers created for the Wii.
2. Aggregation of individual results from the parts of a complex system. The financial crisis is a great example. The interactions of the different market participants aggregated into an economic meltdown.
3. Failure to review policies and procedures. Policies and procedures that may have been useful and necessary at some point in time can cause unintended consequences and delays in business processes if they are not reviewed timely.
4. The problems stated above can create challenges for management in forecasting, mitigating risks and making tradeoffs. The article suggests the following as possible solutions that managers can use.

To improve forecasting methods, managers can:

Complex systems require better risk mitigation that goes beyond traditional methods of risk mitigation. Decision makers should consider the following when improving risk mitigation in a complex system.

1. Vantage point problem. This is a decision maker’s inability to see all the interactions of a complex system.
2. Cognitive limits. A decision maker will have a limited understanding of the consequences of all the actions of a complex system’s various participants.
3. Focusing on one thing can prevent us from seeing others. Those who have done the “invisible gorilla” test appreciate how focusing on the basketball passes prevents most people from seeing the person in the gorilla costume.
4. Rare events. When decision makers fail to consider the “black swan” scenario, they may become overwhelmed and unable to manage the complex system.
   • Drop certain forecasting tools. The article suggests that managers need to drop certain tools that have assumptions that apply to complicated systems and not to complex systems. Tools that assume that features in the system are independent and those that assume that outliers are rare are not very applicable in complex systems.
   • Simulate the behavior of a system. Models that base their design on how a complex system works and models that consider “low-probability but high-impact extremes” are better than those that focus on medians.
   • Use the three types of predictive information. Dividing data into the following buckets makes it easier for managers who find it difficult to forecast outcomes of a complex system.
     1. Lagging:  this includes financial metrics and key performance indicators.
     2. Current:  data of your current situation and performance.
     3. Leading:  data of possible outcomes of the complex system.

The last challenge of making smart tradeoff decisions can be lightened when decision makers:

1. Limit or even eliminate the need for accurate predictions. By nature, complex systems are unpredictable. Therefore, investing in predictions can be fruitless.
2. Separate elements of a complex system from each other to avoid the risk of each affecting one another during an adverse event. Also, redundancies allow better adaptability.
3. Draw on storytelling and counterfactuals. Consider asking the “what if” questions to evaluate the system’s readiness for catastrophic events.
4. Attack a problem from multiple angles using different techniques and assumptions.
   1. Take a real-options approach. Make a lot of small investments that provide options but not obligations. This will help save costs and better manage problems.
   2. Ensure diversity of thought. Bring together teams of differing but complimentary thinkers.

Visit the Harvard Business Review website to obtains this article.