In my previous blog, The Characteristics of Metrics that Matter: Overview, I introduced the subject by listing seven characteristics that determined what makes a good metric. Here, I delve into the first on the list – a good metric must be “relevant.”

Relevance means tight alignment with corporate objectives and there’s no better time to test for that then when you are reviewing corporate objectives or goals. No magic formula here – simply verify that the metric considered is truly derived from one or more objectives. Does a metric’s value or trend let you confidently know whether you missed or met an objective? If you’re monitoring percentage of accepted employment offers, then improving the hiring process, increasing competitiveness of offers, and/or bolstering recruiting follow up should be a part of organizational objectives.

Now that you have confidence that the metric monitors a valid objective, check to avoid imbalance with other metrics or “sub-optimization” when one or more metrics undermine the objectives of others. For instance, you’re hurting profit (objective) if reducing purchasing costs by a particular percentage harms product quality and increases overall costs through unexpectedly excessive scrap and/or rework. You can avoid this in two ways.

First, derive or check existing metrics with respect to all corporate objectives at the same time to understand how those metrics relate to one another and ensure this is done each time objectives are reviewed. Second, combine multiple metrics into an “index” value that drives a larger goal such as customer satisfaction with an order process (the so-called “Perfect Order”). For instance, multiplying percentage on-time product delivery x percentage order completion x percentage deliveries free of damage x percentage shipping that includes an invoice would give an index that considers even competing metrics.

Assessing relevance should not be an “ivory tower” exercise. Survey appropriate individuals up and down the corporate hierarchy to question metric relevance and adjust if needed. Relevant metrics must drive behaviors and their associated outcomes and getting buy-in should confirm valid, potent, and energizing behaviors. Remember that, as you move up the process “food chain,” higher-level metrics must effectively tie back to those at lower levels. The metric related to a high-level corporate goal may not mean much to a clerk processing invoices unless there is a relevant linkage to him/her.

Finally, a test of relevance relates to the ability to understand the total, actual impact of a change in the metric. For instance, if MTTR (mean time to repair/resolution) increases by 10 minutes, can you estimate what that real, total impact will be? MTTR changes can affect staff assignments, customer satisfaction, trouble ticket creation, and other process components. If we become faster or slower, can we calculate how much time/money/resources we save or lose?

If a virus attack causes a server outage, the time to get the server back up and running is not necessarily the actual or only impact. The impact is when full production is brought back up (recovery of lost data, rescheduling of incomplete jobs, etc.). The better you know metric impacts, the easier to judge the relevance of a metric to goals and/or objectives. 

How do you assure and/or measure relevance in metrics? Drop me a note and let’s discuss.

Eric Noack is the Director of Information Delivery at Expert Analytics. His project history spans numerous Fortune 500 and national-to global-scale companies with dozens of projects in accounting/finance, web analytics, budgeting, services, and retail subject areas. 

At Expert Analytics we provide data experts and expert teams to carry out your strategic data initiatives expeditiously and cost-effectively. Our experts are experienced in a wide range of data tools and platforms and are highly credentialed with decades of successful client deliverables, advanced degrees, and certifications.

Our experts help you achieve success with your most strategic initiatives and improve your teams’ performance across a range of projects through mentoring and advanced expertise.

Expert Analytics, based in Dallas, Texas, was created by data warehousing and analytics veterans and thought leaders.  From the beginning, expertise is the company’s culture and is deeply woven into every client delivery. 

www.ExpertAnalytics.com

(800) XPRT137  |  (800) 977-8137