Three Critical Legs to Top Business Performance

January 3, 2017

Efforts to drive top performance in an organization are significantly compromised without an emphasis on and balance among three critical components of a business’s organization:

  • Are workers empowered to make decisions?
  • Are they measured on their performance?
  • Are they incentivized for their performance?

In the first empirical study on the simultaneous interplay among these three important components of organizational structure, researchers found that an organization’s performance is significantly improved when firmly based on a balanced deployment of decision rights, performance measurement, and incentives.

“The study provides unique empirical evidence on the relevance of organizational architecture as a concept to explain organizational performance. We find that organizations with higher delegation, performance measurement, and rewards report better performance.”

Three legs of business performance stool graphic

Mahendra Gupta, former dean and Geraldine J. and Robert L. Virgil Professor of Accounting—along with his coauthors—compare those components to the legs of a three-legged stool. When a company delegates decision rights to its managers, tracks their performance, and ties performance to incentives, the company’s overall performance benefits. Take away one of those legs and performance falters.

Previous studies focused on the benefits of delegating decisions to managers, or on the relationship between performance evaluation systems and incentives, but never all three together. This new study is the first to link these components together and show that as a harmonious ensemble, they form the high performing “organizational architecture.”

In other words, if you want to maximize your firm’s performance, it is not enough to track your managers’ performance. Simply providing managers with incentives is not enough; nor is giving them more latitude in decision making. You must have all three in place for an effective organizational architecture that will enhance performance.

“Arguably, organizational investment of time and resources into establishing performance measures and rewards or formulating the appropriate delegations of authority communicate organizational priorities to which managers are responsive.”

Testing this three-legged stool theory had eluded researchers until now because of a lack of organizational data. Gupta and his coauthors turned to a unique database containing purchasing card information from 586 organizations. Companies obtain purchasing cards from financial institutions and distribute them to certain employees. The purchasing card serves as a charge card that employees use when making purchases from vendors. The company receives and pays a single invoice each month for all cardholder transactions.

Organizations have found that using cards in their purchasing process saves significant time and money, yielding improved organizational efficiency and management of transaction risk. This may explain why 85 percent of large North American companies have adopted purchasing cards over the past 20 years, accounting for more than $300 billion in annual low-dollar purchases across an estimated 574 million transactions, yielding approximately $40 billion in transaction cost savings for these organizations.

Why are purchasing card programs an ideal research setting to test the importance of the three-legged stool theory? Companies that adopt such a program must have a purchasing card administrator—a person who is typically not an upper-level executive. This allows researchers to examine the effects of delegating decision-making to the PCA as well as using a performance measurement system to track the PCA’s performance, and incentives to nudge the PCA’s behavior in the right direction.

To turn each of these concepts into variables they could statistically analyze, the researchers used survey data on the use and management of purchasing cards from a market research firm.

  • The “decision rights” variable measured the extent to which a PCA had influence over card program decisions.
  • The “performance measurement” variable measured whether a performance metric was used and, if so, to what extent.
  • The “incentives” variable measured the effect of purchasing card program performance on the PCA’s pay, job security, and prospects of getting promoted.

The researchers divided the companies into groups based on whether each measure was above or below the median and tested to see which firms performed the best, with two different measures of performance: the percentage of low-value (below $2,500) purchases versus higher-value purchases ($2,500 to $10,000) made by p-card.

In each test, the companies that scored above the median on all three measures—companies that delegated decision-making, used a performance measurement system, and provided the manager with incentives—outperformed the companies with a different organizational structure.

KEY TAKEAWAYS for Managers

  • An organization’s architecture is linked to its performance.
  • A high-performing organizational architecture is a three-legged stool comprised of delegated decision-making, performance measurement, and incentivized people.
  • Having one or two legs of the stool isn’t enough; all three must be present and working in balance for a company to achieve best performance.

The authors have provided powerful empirical evidence in support of their three-legged organizational structure theory with metrics demonstrating that “organizations with higher delegation, and greater emphasis on performance measurement and rewards report better performance.”

Mahendra Gupta presented his research to alumni and friends in the business community on January 11, 2018.

“The Performance Effects of Organizational Architecture”


Mahendra R. Gupta, former dean and Geraldine J. and Robert L. Virgil Professor of accounting and management, Olin Business School, Washington University; Antonio Davila, professor of entrepreneurship and accounting and control, IESE Business School, University of Navarra; and Richard J. Palmer, professor, Harrison College of Business, Southeast Missouri State University