Step 6: Statistical analysis

After normalization, a company (or its service provider) is ready to statistically evaluate the benchmark data. It can calculate quartile break points for each metric or service and quantify the performance gap between the benchmarking organization and its peers. This analysis will drive the formulation of performance improvement initiatives or, in the case of sourcing benchmarking, may serve as a basis for price negotiations. Bar charts can be used to highlight quartile breaks and areas needing improvement in the benchmarking organization. They make the statistical analysis easy to understand and provide a clear basis for setting targets for the benchmarked metrics.

Step 7: Finalization of project deliverables

A service provider's final deliverable for a benchmarking project should be a report providing ranges of performance for metrics in each of the defined services, along with recommendations for changes suggested by the analysis. The report typically also includes backup documentation consisting of a situation overview; a description of the approach to benchmarking; identification of the factors considered in normalization; comments on the adjustment of these factors; a template of the benchmarking framework, which summarizes the services being benchmarked; assumptions underlying the analysis; and the findings of the benchmarking process. Opinions and recommendations for improvement initiatives should be included wherever they are relevant, or wherever differences are stark between the benchmarking organization and its competitors. The process should be fully inclusive, and conclusions and findings should be shared throughout the process. The final report should include no surprises.

Because of the number of variables it entails, custom benchmarking is one of the few reliable methodologies available for validating the metrics and targets within an organization's performance management framework. Having a third party manage the benchmarking process, and analyze the results, helps ensure that the study is done objectively and that it encompasses all relevant factors in making the comparisons.

The unique circumstances in which a company operates make any generic or off-the-shelf benchmark report suspect. No standardized report can fully accommodate the individual situation within a given company. As pressure on business performance continues to increase, companies will find it increasingly crucial to be able to recognize how they are doing. Truly comparable benchmark data will be the only way to appropriately adjust their activities to changing market factors. Customized benchmarking is one of the key tools any business can use to ensure sustained long-term performance.

John Hall (john.hall@paconsulting.com) is a principal consultant with PA Consulting Group, specializing in performance and asset management and organization design.

Yann Pastor (yann.pastor@paconsulting.com) is a principal consultant with PA Consulting Group, specializing in strategy and value-based management.

Selecting the Right Metrics

Managers struggling to choose metrics for their organization should consider a value-based approach. Such an approach helps focus the company's performance management processes on the activities that actually impact stakeholder value. There are as many facets of performance as there are stakeholders in the business, so it's useful to sort candidate metrics by stakeholder group. For instance, an organization might categorize its metrics into four buckets: value to shareholders, which would include measures of profitability, cost, and revenue; value to customers, including measures of customer satisfaction or service-level performance; value to employees, such as OSHA metrics or a satisfaction index; and operational excellence — e.g., quality, compliance, productivity.

To ensure that you select the right metrics, follow these rules when designing your performance management framework:

1. The metrics should be mutually exclusive and collectively exhaustive. They should neither repeat between scorecards nor overlap.

2. Each lower-level metric should align with a key companywide metric. We use a technique called value-driver analysis to map metrics in a hierarchical “value driver tree.” Especially when corporate metrics are financial in nature, lower-level metrics — both financial and operational — should link to them mathematically in a clear hierarchy.

3. All metrics should be easy to understand, both by the employees who are accountable for them and by those responsible for reporting on them.

4. Keep the number of metrics to an absolute minimum. If two cover the same area, remove one or look for an alternative that can replace both.

5. Where benchmark data is available, align metrics with benchmarks to ensure that external data is an integral part of the target-setting process.