Knowing your Operating Model

Don’t get off track on your way to organizational transformation

The idea of an “organizational operating model” is the central concept discussed by Dr. Jeanne Ross, Peter Weill, and David C. Robertson in their seminal work, Enterprise Architecture as Strategy: Creating a Foundation for Business Execution. They contend that for companies to succeed, they must understand their operating model: in other words, the degree to which they must standardize business processes and/or integrate data to produce optimized business outcomes. Higher standardization of processes is typically required where efficiency and predictability are the primary factors in delivery of the organization’s products or services. For instance, if an organization wants to present the same face to customers in diverse locations, or capitalize on efficiencies of scale in purchasing, a high degree of standardization is required. However, the higher the level of standardization, the greater the cost to the overall organization in terms of flexibility and the ability to independently innovate. I have found that often the only words that get heard with regard to the enterprise architecture are “higher level of standardization” and “higher degree of integration.”

As a manager, the information insight into your operating model should drive governance as well as your EA and planning programs. In fact, the operating model should drive every other decision you will make because it will help you understand the federation and spans of control within the organization; to the point where you can determine at what level decisions should be made regarding specific types of management inputs. Particularly for large, complex and/or federated organizations, this is one of the major sticking points that prevent the organization from getting value from their enterprise architecture, governance, and planning efforts. A large diversified conglomerate simply should not be looking to standardize and integrate at the enterprise level in the same fashion that a tightly focused technology oriented or manufacturing company should be, even at comparable sizes. The same advice follows in the public sector. Agencies with tightly focused missions like the IRS can have a much more standardized and integrated planning and governance configuration than an organization like the USDA, which includes organizations dedicated to emergency response, research and development, financial management (loans & grants), etc.

That isn’t to say that there is not a use for enterprise architecture, governance or other planning disciplines within organizations that are using a diversification model (low standardization, low integration), it is simply that the function of these efforts may be more focused on reducing core infrastructure costs, managing licensing, supporting capital planning, and ensuring a technology backbone is in place to support the business. This is in contrast to the unification model (high standardization, high integration) that may drive a ROI in a highly standardized and integrated organization, like an online retailer, where massive economies of scale and top line value may be gained by tightly controlling decisions and resources. Obviously, there are degrees to all of this and I haven’t even mentioned coordination or replication models, but my point is that there is real value in examining your organizations operating model as an input to how you build your EA, governance and planning efforts. In the end, these management disciplines developed with the intention of facilitating decision support and control need to be tailored to the environments in which they exist. An examination of the operating model will help you plan at the right level and delegate in a manner that reduces the tax that many business units perceive is being placed on them when “corporate” comes asking for data or begins to exert influence on buying decisions. Getting value from these types of programs once there is a perception that the governance or planning process is increasing response times, lowering performance and adding to the cost to execute is nearly impossible.