Impact of Formal Controls on Client Satisfaction and Profitability in Strategic Outsourcing Contracts

Abstract

The reach and impact of outsourcing has grown rapidly to include a variety of strategic objectives. Significant issues of cooperation and coordination in strategic outsourcing necessitate investments in formal controls that specify and monitor execution of the outsourced task to mitigate unforeseen contingencies and improve predictability in the attainment of desirable goals. In this study, we investigate the nature of formal controls in strategic outsourcing contracts and their impact on client satisfaction and financial performance. Specifically, using rich field data on 390 strategic outsourcing contracts, we examine the differential impact of output controls, activity controls, and capability controls on client satisfaction and vendor profitability. We find that activity and capability controls are positively associated with client satisfaction and profitability; in contrast, output controls differentially impact satisfaction and profitability, reflecting potential tradeoffs between the two outcomes. Our results, in addition to contributing to the research in control theory, provide actionable insights for technology vendors into the appropriate strategy and tactics required to compete efficiently and effectively in strategic services markets.

Key words and phrases:

• outsourcing
• strategic outsourcing
• formal controls
• client satisfaction
• profitability
• IT outsourcing
• outsourcing controls

Acknowledgements

We thank the Srini Raju Center for Information Technology and the Networked Economy at the Indian School of Business for helping us access the data used in this study. For helpful comments on earlier drafts, we thank Kannan Srikanth, Jens Dibberns, Vijay Gurbaxani, numerous participants at the 2013 International Conference on Information Systems, and seminar participants at Georgia Tech and Nanyang Business School at Nanyang Technological University. We are also grateful to the review team for their meticulous and detailed comments that helped improve this paper. Finally, we thank senior management at our research site for their generosity in sharing this dataset and for their time and effort in clarifying our numerous questions. Any errors remain ours.

Supplemental Material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

¹ See the online Appendix for an example of a strategic outsourcing contract between EDS and General Motors.

² We checked for the presence of common-method bias through Harman’s single-factor test [72]. All of the variables in our study were simultaneously subject to an exploratory factor analysis, and the results of the unrotated factor solution were examined. The absence of a single factor that explained a significant amount of variance in the data suggested that common method bias did not likely impact survey responses. In addition, we used Cronbach’s alpha to check the reliability of constructs.

In addition, though the nature of the data does not allow us to use the Multitrait-Multimethod Matrix [13], we estimate a structural equation model to establish the construct validity of the measures. High correlations between satisfaction scores and the latent variable CSAT_Factor (all factors load to a single factor, see online Appendix Table A.1) and low correlations between this variable and other subjective measures such as financial performance (FIN_PERF), business controls (BIZ_CTRLS), and service quality (SQUAL) (all < 0.40), manifest both convergent and discriminant validity. For interested readers, the full SEM estimates are available from the authors upon request.

³ The details of this factor analysis are presented in the online Appendix (Table A.1) and show that each variable loads highly on this single factor (CSAT_Factor). This factor was highly correlated with CSAT (r = 0.9277, p < 0.0000). As a robustness check, we estimated our model with CSAT_Factor instead of CSAT. Our findings remain consistent and are available from authors upon request. However, CSAT is a simpler measure to interpret; henceforth, we use CSAT in our empirical analysis and discussion.

⁴ It is possible that the error terms are indeed correlated across the two equations. To evaluate this, we estimate a Seemingly Unrelated Regression (SUR) model (see Equations A.1 and A.2 in the online Appendix); our estimates from the SUR model indicate that the correlation between the error terms in the two equations is insignificant. The Breusch-Pagan test of independence for 1) Model specification with interactions with total competencies, χ²(1) = 1.494, p = 0.2216, and 2) Model specification with interactions with total service lines, χ²(1) = 2.873, p = 0.0901. Therefore, the hierarchical Bayesian model (Equations 1(1) $\begin{array}{rl}& CSA{T}_{i,t}={\alpha }_{0i}+{\alpha }_{1i}.LEA{N}_{i,t}+{\alpha }_{2i}.BIZ\mathrm{\_}CTR{L}_{i,t}+{\alpha }_{3i}.DEVM\mathrm{\_}CT{R}_{i,t}+{\alpha }_{4i}.EM\mathrm{\_}CT{R}_{i,t}+\\ & {\alpha }_{5i}.SLO{S}_{i,t}+{\alpha }_{6i}.SLA{S}_{i,t}+{\alpha }_{7i}.SQUA{L}_{i,t}+{\alpha }_{8i}.TENUR{E}_{i,t}+{\alpha }_{9i}.TC{V}_i+\\ & {\alpha }_{10i}.CONTROL{S}_{i,t}+\epsilon 1i,t\end{array}$(1) -4(4) ${\beta }_i={\delta }_{20}+{\delta }_{21}.TOTAL\mathrm{\_}COMPTNC{S}_i+{\delta }_{22}.TOTAL\mathrm{\_}SVC\mathrm{\_}LINE{S}_i+v_{2i}$(4) ) assumes independence between the two outcome equations.

⁵ Estimation using Bayesian inference techniques can be more parsimonious in their data requirements as the estimation procedure is able to partially pool data across observations and thus present more information that can help estimate the individual-specific parameters [3, 80]. Allenby and Rossi [3] offer more discussion on the appeal and constraints of Bayesian inference.

⁶ We also conducted two diagnostics tests—Geweke convergence test [30], and Heidelberger and Welch’s stationarity test [42]—to check whether the estimation reached convergence, and found adequate convergence. More details on the MCMC simulation are available from the authors on request.

Additional information

Notes on contributors

Nishtha Langer

Nishtha Langer ([email protected]; corresponding author) is a member of the Information Systems, Operations, and Marketing faculty at the Lally School of Management, Rensselaer Polytechnic Institute. She holds a Ph.D. in Management (with specialization in Information Systems) from the Tepper School of Business, Carnegie Mellon University. Her research examines organizational and management aspects of key IT resources, both IT capital (process and technology) as well as IT labor (human capital). Dr. Langer’s work has been published in such journals as Management Science, Information Systems Research, and MIS Quarterly, as well as in leading conference proceedings and edited book chapters, and widely presented to both academic and industry audiences.

Deepa Mani

Deepa Mani ([email protected]) is a member of the Information Systems faculty and Research Fellow at the Indian School of Business (ISB). She is also the Executive Director of the Srini Raju Center for Technology and the Networked Economy at ISB. She received her doctorate degree from the University of Texas at Austin. Dr. Mani’s research interests are at the intersection of technology, organization, and society. She is studying the organization of IT and IT-enabled business functions in firms, and assessment of the business value of a broad class of IT decisions and interventions. Her work has been published in leading academic journals, numerous refereed conference proceedings, and edited book chapters, and featured in popular media outlets. She serves as an Associate Editor at Management Science and Information Systems Research.