Corporate decision-making is supposed to be done by committee of an executive board, representing all parts of the core business functions in an organization:
finance, production, manufacturing, marketing, human resources, IT, sales. In many cases, the CEO will exercise authoritarian decision-making based on perceptions or a solitary sense of purposeful guidance. In some cases, CEOs look externally to secure endorsement for their ideas or initiatives. In other cases, a more systematic approach is used to generate sound decisions.
In the field of decision science, Marshall, Kneale, & Oliver (1995) state that there are six concepts that factor into the arrival of a decision (see Figure 1).
Every decision has an entering objective, for example, to lower costs of MIS support, or, to provide speedier response to customer support calls. Every decision being considered has certain characteristics. For example, costs of MIS include equipment costs and staffing and lower consumption costs, as well as space costs. Costs of providing customer service include cost of the systems and staff hourly labor to answer calls.
Next, attributes of each characteristic are measurable ways that we can evaluate the costs of the characteristics: for example, dollars and cents, or time in minutes and hours. Quality of customer service support may be expressed in number of complaints or number of hang-ups before an operator answers the call. Next, continuing the aforementioned two examples, each attribute also has an associated criterion that is the level by which we measure improvement, or satisfactory performance or other levels of success.
For costs of an MIS, the criterion may be to lower the dollar costs by 20% per year, without a negative change in the level of service, as evidenced by a change
in the number of help desk calls. For customer service, the criterion may be an improvement in customer perception of our support as evidenced by the number of positive responses to a customer survey done every three months.
Next, we have trade-offs which are exchanges of attributes to achieve a benefit to the organization. For example, we may accept higher costs (in dollars) for a lower rate of customer complaints (in number of negative responses). Lastly, all decisions are framed with some constraints that we must understand in order evaluate the feasibility of a decision. Constraints to lowering costs for an MIS might include fixed costs already incurred for investments made, or for mainte-nance contracts entered into. Constraints to improving the quality of customer service support may be the comparative level of quality of the underlying products we have sold. In some cases, a company will bring in outside support to help negotiate a particular decision, in order to take the perception of close association with a decision away from senior management, or to help the decision appear more objective.
Fisher Scientific is an advanced instrumentation company. Consultants were used to endorse outsourcing IT with mixed results. In the late 1990s, senior management created an independent, but internal Operational Strategies Group (OSG) to explore untapped areas for increased profitability. The OSG project reported directly to the CIO. The group enlisted external consultants to give further weight and credence to their steering committee studies and recommen-dations. OSG worked on projects ranging from enhancement of data warehouse capabilities to determining the causes of profit margin erosion. The group even recommended restructure of the sales force to better fit market conditions. In this instance the CIO also used consultants to get an outside second opinion. The consultants recommended to OSG and the CIO that the firm outsource much of the existing IT department to cut costs and to bring a fresh view of the architecture.
The internal changes implemented by the OSG were successful. The IT outsourcing was pilot-tested but did not achieve the success of the internal Figure 1. Decision model components
1. Objectives 2. Characteristics 3. Attributes 4. Criteria 5. Trade-Offs 6. Constraints
projects. While using independent consultants is often useful for fresh perspec-tives and innovative insights, sometimes all the factors of a good decision are not taken into account. In this case, the complexities of Fisher Scientifics’ legacy systems was too intricate for outside consultants to comprehend, and the outsourcer was unable to completely handle migration of the MIS services. Only internal IT staff with many years of experience with the MIS were able to handle them.
It is evident in terms of decision models, that when one decision works for a particular set of projects, the same decision model criteria applied to a slightly different operating or MIS model may not yield the same results. In this case, constraints were not accurately understood, and the attribute of lowered cost alone was not sufficient as a criterion for successful outsourcing MIS. There-fore, every decision should be evaluated for both infrastructure as well as a detailed understanding of what currently goes into maintaining an MIS. When the parameters for a current maintenance activity require historical perspective and specialized expertise, it may be dangerous to try and outsource that work.
Alternatively, if the organization could have rather ported the MIS to an outsource provider’s newer platform, this may have proved less risky than assuming all would continue as in the past.
Another area where consultants are often used to drive a decision model to outsourcing is in politically sensitive decisions. As firms increasingly find public outcries and furor over loss of thousands of jobs due to closure of MIS support offices, they find that using external parties to convey these recommendations makes them more palatable. Consider the point that many outsourcing firms are no longer publicizing their mega-outsourcing deals for fear of backlash.
As the boardrooms of large corporations have come to recognize both the large budgetary impact of IT decisions, as well as the strategic advantage that can be wrought by effective uses of IT for e-commerce and new market penetration, corporate executives are paying closer attention to how IT is leveraged for strategic and tactical advantage. In the past, senior executives have often turned to the chief financial officer to analyze trends and to provide quantitative guidance as to productivity, financial results, and for forecasting sales and profits. The chief information officer is now standing at a par with the financial executives. It is no surprise that the marriage of IT with finance represents such a strong decision model for business. By using financial analysis models to support IT proposals, the CIO has a better chance of convincing their peers in the board room of the wisdom of outsourcing arrangements. This is discussed further in the section on decision-making models.
One outsourcing deal occurred well before the advent of millennium scares and enterprise resource planning systems and Internet e-commerce MIS. CIO Magazine (1999) discusses how Kodak used multiple outsourcing vendor
arrangements as part of its IT strategy as early as 1989. Kodak hired multiple vendors to handle data center operations, telecommunications, and desktop support. Kodak’s decision to outsource to multiple best-of-breed vendors was considered risky due to the possibility of loss of control to its customer and production systems. However, by using vendors to provide basic IT resource management services, within a year, Kodak’s IT capital costs dropped 95%, PC support costs dropped 10%, and mainframe operations’ costs dropped by 15%.
Kodak continues to outsource these support services, and continues to save IT support costs, through renewal of the original deals with many of the original vendors as well as with new vendors (CIO, 1999). The obvious decision model attributes for these multiple arrangements were cost related. Kodak, however, did not ignore quality as an attribute and customer perception as criteria for the quality attribute. In selecting its outsourcing partners, Kodak uses reputation as a means of gauging the ability of its outsourcers as equal criteria for selection, equal to lower cost and greater efficiency of operations. In another sad case, however, cost alone was the attribute of importance. Level of cost in comparison to price was the criteria — and the lower the better for Armstrong.
In 2003, Armstrong World Industries outsourced much of its computer systems applications support in a joint deal with Computer Sciences Corporation (CSC) and Satyam, an off-shore outsourcing firm based in India. The deal was valued at $2.6M annually for a term of three years. Forty-three long-time Armstrong employees were fired as a result of the deal although they were given incentives to stay for three months to transfer knowledge to the outsourcer (Applegate, 2003).
The structure of the deal was unusual in that Armstrong would only have contact with CSC as the intermediary. CSC was to gather work requirements and communicate these to the third party outsourcer, Satyam, for development. In essence, CSC outsourced the most important part of the outsourcing activity, the programming, to another outsourcer. This needlessly complex arrangement only served to add costs and create a communication chain where messages from Armstrong were often misinterpreted by the time they reached the Satyam programmers.
The arrangement began in February of 2003 and productivity dropped by 60%, as measured in work orders closed per employee. While productivity was dropping, the rate of programming requests increased. The resulting “bottle-neck” caused nearly all IT work to halt. So much administrative paperwork was involved in authorizing work that the programmers in India often spent entire days with nothing to do awaiting authorization, despite a growing work backlog at CSC and the client. Despite the fact that CSC increased staffing and hired back as many former Armstrong programmers as possible, the backlog contin-ued to grow. Less than nine months after the agreement was signed, Armstrong
terminated the agreement, citing CSC’s inability to meet its target performance metrics. The primary cause of the outsourcing failure was underestimating the amount of business and system knowledge that was lost when Armstrong immediately fired the 43 people that had been supporting the system. The criteria for successful productivity numbers, work orders closed, was artificial, an ineffective bellwether of success for Armstrong. Communication breakdowns between Armstrong, CSC, and Satyam also doomed the effort.
To further exacerbate the arrangement, security of data also became an attribute that was not well managed in the outsourcing arrangement. Many corporations will still not consider outsourcing their security. In fact, an analysis of the Armstrong case points to the need for even more knowledgeable in-house security and disaster recovery experts to deal with the complexities of a global, inter-networked, outsourced environment. The internal security staff actually increased from four to five people to handle the increased work from the new outsourcing arrangement.
It is clear that some outsourcing deals work well for all parties while others are destined for failure. Good communication, comprehensive metrics, excellent knowledge transfer, and common sense are all key ingredients in making a successful outsourcing deal. The place to begin is in defining the objectives and planning the structure of the outsourcing arrangement well before concluding the contract.