ENTERPRISE SYSTEMS DEPLOYMENT: PAYBACK TIME

by Mark Cotteleer, Senior Consultant, Cutter Consortium

Over the past several years, there has been debate in the academic and industry research communities over the payback of enterprise-level information technologies like enterprise resource planning (ERP) systems. Some have suggested that there is no payback for these investments; that they represent infrastructure investments and are strategic in nature (i.e., we need to pursue them in order to keep up with the competition). Performance-oriented managers should recoil at such logic.

The "no payback" argument is an inappropriate response to the issue. There must be a return, or the firm should not invest. Some have used this line of thinking, however, to avoid the effort of doing a good job with the business case up front and with the post-implementation evaluation following the deployment. The response may be understandable, however, given the extended time horizon that may need consideration for the evaluation of an enterprise systems investment.

A key issue that managers must account for is that a lengthy "adjustment period" may follow implementation of these systems, during which the real value of the implementation may not be apparent. An adjustment period is the amount of time it takes for a firm to fully adopt, and learn to use, a new technology. During the adjustment period, performance may show no improvement -- or may even decline. I thought it would be helpful to consider the information that is out there with regard to adjustment periods following enterprise systems deployment and the influence they have on the ultimate payback of enterprise systems.

Research completed by industry consultants takes a long view of the payback period required to justify enterprise systems implementation. For example, in 1999, one well-known firm offered a figure that suggested paybacks in the range of 30 months. Frequent published accounts suggest that adjustment periods lasting a year or more (followed by eventual performance improvement) are driving these extended paybacks.

Rigorous research in the field of enterprise systems is only now beginning to emerge. What has been completed supports the long view as well. In multiple cases involving my research and the research of others, findings suggest that an adjustment period may follow deployment. During this period, performance may decline prior to improving (depending on how you measure it).

For example, one study looked at two operational measures: on-time shipping and lead-time performance at a single company. On-time shipping actually got worse immediately following deployment of the enterprise system, but it soon began a steady improvement pattern that left performance in this dimension significantly better than its predeployment levels.

Lead time showed a mixed picture, with a decline in performance for some product lines but not for others. There was also a delayed adjustment period that occurred a few months following deployment. In all cases, it took approximately four to six months for post-deployment performance to surpass prior performance.

These findings are echoed somewhat in my research. I have also looked at lead times, this time in the context of multiple sites in a single firm. My findings show immediate improvement for the firm that I studied. However, there is an adjustment period during which the firm "gave back" some of its initial gains before getting better over the long haul. This adjustment period lasted from nine to 11 months at the sites I studied. Drivers of the adjustment period length and intensity have to do with the firm's approach to inventory management, end-user support, and control over local adaptation of the technology and business process.

The concept of an adjustment period is also captured in multi-firm studies. One recent study showed that across a wide spectrum of SAP implementations, initial gains in firm-level measures such as return on assets and "value add" fall in the period immediately following deployment. These gains still exist, but they seem to shrink a bit. This study did not look at a long enough time period to assess whether these factors eventually grew again, but I suspect that they will. The finding across firms is consistent with the single-firm findings that I uncovered.

The implication of these adjustment periods for the payback of enterprise systems is that payback can be substantially delayed as the firm bears the cost of performance stabilization, or it may even decline following deployment. Following the adjustment period, the firm needs an extended period of improved performance in order to recoup both the initial investment and any potential costs of the adjustment period. When considering the value or payback of a potential (or current) enterprise systems investment, managers should do the following:

• Take the long view for these kinds of investments. Make sure you account for the possibility that performance benefits may be realized only after an adjustment period during which the firm learns to use new technology and process.

• Make sure that when you evaluate enterprise systems investments, you do so with a combination of evaluative measures, some of which are robust to delays in the realization of benefits. Measures such as return on investment and net present value will help to appropriately value long-run benefits achieved from the implementation.

• Remember that, in the end, payback will depend on how value is counted and how effectively the implementation occurs. Focus on the measures that are strategically important to your firm, and do a great job with those. Worry less about unimportant measures. When you know what to measure, you can focus on preparing the relevant areas of the firm to receive and use new technology and process.

-- Mark Cotteleer, Senior Consultant, Cutter Consortium