• Strategy
  • McKinsey & Co.

Managing Your Organization by the Evidence

An organization is much more likely to improve its current performance and underlying health by using a combination of practices rather than any one of them alone, new McKinsey research finds.

Furthermore, companies can’t ignore their organizational past. Take one with a rich history of national operating units reflecting different local cultures. When the company decided to merge into a pan-European organization, its future direction was set from the top down, with limited consensus building; detailed individual performance contracts defined accountability, and local leaders were moved to a central hub. The decision to bypass local cultures and to emphasize the performance of individuals rather than teams or units created a major fracture in the new organization. Employees felt unclear about its direction, performance contracts failed to take hold, and both its performance and health suffered.

Most managers are notoriously subjective, prone to manage by anecdote, quick to adopt best practices, and fond of big, visible initiatives. But the evidence from McKinsey’s database suggests that a company’s performance and underlying health are much more likely to be improved by a combination of complementary practices—especially those that provide for clear accountability, help set goals and priorities, and encourage a high-performing corporate culture. Top managers would be wise to base their actions on this evidence of proven success and not on prevailing wisdom and myths, however seductive.

The Data Behind the Findings

The data supporting this article’s conclusions come from McKinsey’s Performance Leadership Survey, an in-depth questionnaire explicitly designed to explore an organization’s effectiveness. Set up in 2002, the database contains information from almost 400 discrete business units of 231 global businesses in all major regions and industry sectors. More than 115,000 individual managers and employees have participated.

Our survey captures two distinct but related aspects of performance—outcomes and practices. The questions about outcomes probe a company’s effectiveness at managing nine organizational elements: direction, leadership, environment and values, accountability, coordination and control, capabilities, motivation, external orientation, and innovation. The questions about practices help to identify the way companies try to achieve these outcomes.

To take accountability as an example, the outcome measured is the extent to which an organization’s people know what they are accountable for doing. The practices creating a sense of accountability range from the design of structures and roles to performance contracts to “consequence systems” (rewards and penalties) to implicit agreements. Any combination of these can in theory achieve the desired outcome.

As for capabilities, the effective outcome is an organization with the right skills and talent to support its strategy and competitive advantage. To that end, a company can use a variety of practices—hiring exclusively at entry level and nurturing talent internally, hiring experienced people, or resorting to contractual and temporary solutions.

Regardless of a company’s choice, the definition of an effective outcome remains the same.

Our analysis of the data focused on the relationship between these variables—outcomes and practices. Two questions particularly engaged our attention. The first was determining the likelihood that a particular practice will deliver a distinctive outcome (meaning that the company using it ranks in the top 25 percent of the database). The other was to find out which combinations of distinctive practices have the greatest chance of generating distinctive outcomes. We found many examples of practices that, by themselves, had limited value but become highly effective when paired with a complementary practice; the combination of clear roles and performance contracts is a good example.

Discuss

Your email address will not be published. Required fields are marked *