Do ES Packages Drive Convergence?

As companies adapt to tough economic times, many hang their hopes on benefits of enterprise-level software (ES) packages such as SAP, seeking to lock in the “best practices” they perceive these products are built upon. However, critics question whether these pricey software packages can deliver sustained performance improvement. A recent working paper, “An Empirical Study of Operational Performance Convergence Following Enterprise-IT Implementation,” presents a case study of Tristen Corp., a pseudonym for a $4 billion U.S.-based manufacturer of computer components. The author, Mark J. Cotteleer, a postdoctoral fellow at Harvard Business School, documents the implementation of an enterprise-resource-planning (ERP) software package in three of the company's divisions — North America, Europe and Asia. Tristen, which deems the number of days between order placement and fulfillment to be a key indicator of competitive success, implemented its ES package in an attempt to standardize lead time across its divisions. The company also sought to standardize levels of service and improve customer satisfaction.

Cotteleer collected lead-time data covering approximately 113,000 orders from 12 months before through 24 months after ERP deployment. He then looked at the results to determine if the performance of each had improved and if they had converged toward a common, improved lead time.

The data show an initial improvement in lead time across divisions. For example, North America improved dramatically, reducing lead time from 50.6 days four months prior to ES deployment to 29.5 days one month after. Europe and Asia, which already had shorter lead times than North America, sustained smaller initial improvements. This pattern of lead-time improvement led to a sharp convergence in performance across the regions.

Following initial improvements, performance began a steady decline across all regions during the first year after ES deployment. These significant declines did not dip below the levels initially attained after the ES deployment. However, near the end of the first year, this trend reversed, as each region embarked on an extended period of performance improvement through the end of the study period. Cotteleer points to inventory management, personnel and systems adaptation as influencing these trends.

Following the initial convergence across the business units, performance once again began to diverge. Cotteleer identifies factors related to the ES deployment effort, implementation-team composition and differences in the availability of postdeployment support as potential causes of the ultimate divergence in performance.

Cotteleer concludes that managers should not assume that ES implementation will standardize operation performance across the board. Since lead-time convergence for a single company implementing the same ES package was unsustainable, Cotteleer argues that improvements or declines in operational performance may go beyond a single technology to a host of other factors.

For more information, contact the author at mark@cotteleer.com.