How to Increase the Reliability of Your IT Infrastructure Using Predictive Analytics REGISTER >
Each of these valuation methods assumes that the data needed to generate metrics is being gathered and that some baseline knowledge of the current system is available. Baseline metrics are needed both to support the initial ROI justification for a proposed system and to allow for its ongoing evaluation.
In the case of an e-business implementation that will affect customer support, for instance, metrics assessing the customer experience need to be established before the system is put in place. Without setting a baseline that includes such criteria as customer satisfaction, customer retention rates and new-customer adoption rates, for example, it will be difficult to set realistic expectations for a new systemmuch less gauge the actual benefits.
"It was important for us to have clear expectations going in," says Liz Lloyd, manager of information systems for North Shore. "We brought in an outside firm and spent a couple of months working to understand the scope of what had to be done before we chose a vendor. This first round of analysis helped produce reasonable estimates rather than guesses."
Establishing metrics at North Shore, however, required a lot of estimation and validation of data that didn't previously exist. At North Shore, says McHarg, now a partner with M3, a consultancy in Vancouver, B.C., "We had to take the statistics deeper into the organization than we had originally anticipated. But that reality check with salespeople, relationship managers and others was critical."
McHarg is a strong advocate of metrics that let project owners set precise implementation targets. "When you get going on an e-business implementation, suddenly everything competes for importance," she says. "Building out the baseline and benchmark targets with Pivotal beforehand was invaluable to me in highlighting those key opportunities and focusing our development."
E-business systems can often provide the data that is needed to measure its impact on the company. Analytics that cut across CRM, sales-force automation and e-marketing systems can help track the metrics that matter, such as costs to acquire a customer and average revenue per customer.
It used to be that the customer's experience, for instance, could only be captured through a survey or by aggregating such basic statistics as call-center queue times. An advanced CRM system with analytical capabilities can provide detailed information on individual customers, including their use of customer service channels, and which parts of the Web site they use, and how. In the process, larger patterns in sales, usage and satisfaction will emerge. While the primary purpose of such tools is to help department and line-of-business managers make sales and marketing decisions, it can also help evaluate the e-business investment's contribution to business performance.
Christine Cournoyer, CIO and senior vice president of global e-business transformation at Lotus Development Corp., a division of IBM, focuses on what she calls process metrics and results metrics when evaluating e-business systems. The revenue Lotus derives from e-business is a result metric. But it can be just as important to measure the value of the processes that led to that improvement, such as the number of hits to a Web marketing campaign, how many and which pages were viewed, how many visits generated leads and how many led to sales.
The first step is to understand the e-business project's business objectives and use ROI approaches that help build justification. Then, establish process and outcome metrics for identifying whether those objectives are being met. The ultimate goal is to leverage that e-business investment into something that fundamentally changes the company's ability to compete and succeed.