Profiting from knowledge management

Introduction

Systems and processes designed to aid performance fall under the instrumentation category of Gilbert’s behavior engineering model. Often, the value of these tools depends not only on the work situation but also on the attributes of workers they are intended to help. Research uses hierarchical linear modeling (HLM) to examine the impact of time and worker experience on the effectiveness of a knowledge management systems designed to improve sales.

Article

Ko, D., &  Dennis, A. R. (2011). Profiting from knowledge management: The impact of time and experience. Information Systems Research, 22(1), 134-152. doi: http://dx.doi.org/10.1287/isre.1090.0247

Background

Many firms seek to leverage tacit and explicit organizational knowledge to enhance performance through the use of knowledge management systems (KMS). KMS systems take two forms: a codification system, in which knowledge is coded and stored as reusable documents in a central database, or a personalized system, in which knowledge is shared through person-to-person contact on an ad hoc basis. The codification, or people-to-document, system is more formal and requires a substantial investment to vet and organize information in the database. Because of the expense, organizations typically follow the 80-20 rule by selecting the most common 20% of its knowledge likely to be of value to 80% of users.

Previous research focused on whether investing in a codified KMS is worth the cost. Some results indicate that use of a codified KMS actually impairs performance because it diverts worker time away from tasks. However, such research evaluated KMS use and outputs shortly after the KMS was established. The authors of this study posit that the benefits of using a KMS emerge only after some time has passed and workers were able to integrate the knowledge gained into their daily performance. They decided to explore the relationship of KMS use and performance over a two-year timeframe for employees with varying degrees of job experience.

Research

The research focused on KMS usage data and sales performance of 2,154 sales representatives in the U.S. across six divisions of an international pharmaceutical firm. They collected data over a 24-month period (January 2002 through December 2003) and analyzed it to test three hypotheses:

H1. Benefits from the use of a codification-based KMS increase over time.

H2A. Experience will moderate the relationship between the use of a codification-based KMS and the speed at which benefits from use appear such that more experienced knowledge workers will derive benefits faster than less experienced workers.

H2B. The differential performance benefits between more and less experienced knowledge workers from the use of a codification-based KMS will converge over time.

Data used for the study were quantitative and triangulated from three independent sources: KMS usage data from the company’s KMS group; experience data from the company’s personnel management system, and sales performance data managed by a third-party outside contractor. The dependent variable was monthly sales performance defined as the individual’s percent of sales quota achieved. Independent variables were individuals’ KMS use and experience in the current position indicated by number of months.

Prior year’s sales performance and company division were designated control variables based on the company’s KMS group identifying these variables as factors affecting performance. Experience was found to be non-normally distributed, so a log transformation was used on the data prior to analysis. No reliability tests were run since the data were collected from three separate departments and systemic bias that could impact results was unlikely.

To avoid Type I errors, the researchers used hierarchical linear modeling (HLM) rather than traditional regression. The intraclass correlation (ICC) of 0.288 indicates that the use of HLM is appropriate.

Findings and Implications

The researchers compared data on KMS usage to performance over time. Differences in correlation coefficients were significant up to three months. Thus, H1 is supported, although the benefits from KMS use level off and fail to increase after three months. H2A argued that more experienced individuals would show immediate performance benefits from KMS use and that less-experienced individuals would take longer to realize similar gains. H2B argued that over time , the performance differences between the more experienced and less experienced workers would disappear. Researchers examined the significance of experience as a moderator of the relationship between use and performance. They conducted tests for current performance and past performance at one, two, and three months. Models for the current month and one and two months prior show significant positive coefficients on experience, whereas the model for three months ago does not. More experienced sales reps gained greater performance benefits from KMS use than did less experienced reps for three of the four months studied, but over time, those differences disappeared. Thus, H2A and H2B are supported.

Findings suggest that more experienced knowledge workers see faster gains from using KMS because they can quickly recognize, contextualize, and adapt new knowledge to their work practices. Researchers believe that mental models of those with more experience make them better at absorbing and customizing new knowledge to meet their needs. Less experienced workers take much longer to appropriately integrate KMS knowledge into their work practices.

Questions for OPWL-N Members

Is your organization using a KMS? Would a KMS be a worthwhile intervention for your workplace? Which type of KMS would be more appropriate for your workplace: people-to-document or people-to-people? Why?

Workplace Oriented Research Central (WORC)
Prepared by OPWL Graduate Assistant, Susan Virgilio
Directed by OPWL Professor, Yonnie Chyung
Posted on November 5, 2012