Three ways to produce metrics that really matter to your business

We now have data available and digitized from every corner of the business -- from store-by-store sales to employee attendance to lunchroom sandwich sales. However, within all this information, only a handful of data really matters to the advancement and growth of the organization. The question is, what data really matters?

Financial data, of course, is paramount. While we have all the data, many decision makers don't know what to do with it -- or even understand it. A few months back, we posted research that showed an alarming degree of "financial illiteracy" among executives, particularly when it comes to deciphering balance sheets and other statements.

There's also much that can be done on the report generation side of the equation to increase people's understanding of the data they need to know to make effective decisions. One thing that urgently needs to be addressed is filtering out all the noisy, essentially useless information from the few nuggets that really matter.

Eric Ries, writing in the Harvard Business Review blog, cautions against relying too much on what he calls "vanity metrics" to assess the state of your business. Vanity metrics are one issues -- numbers that sound good, but do little to help in pending decisions. Pageviews from a Web site is one example, "billions and billions served" is another.

The other issue is over-reliance on analytics tool-generated numbers that may not have relevance to the growth of the business.

Metrics in reporting should be actionable, accessible, and auditable:

• Actionable: The information should be replicable, and put into context. It's nice to know that in January, 100,000 units of your product were sold to 20-29 year olds as a result of a set of promotions. But what were the results of one style of promotion versus another? As Ries puts it, "techniques like split-testing, where discrete groups of customers see different versions of the product, are the gold-standard here." This creates actionable metrics.
• Accessible: "Most data warehousing systems provide reports that are too complicated to read and take too long to generate," Ries points out. Reports need to be easily accessible, meaning that decision makers should be able to generate the reports themselves. They should also be easily understandable. To add to Ries' thoughts in this regard, visual "Google Maps"-style mashup applications really hit home in terms of understandability. For example, imagine displaying a map that illuminates, at a glance, where your top-grossing retail locations are.
• Auditable: Credibility is another key issue in today's systems. It's important to work closely with the people generating data at the source to assure that it is providing the most accurate picture possible. "It's important that skeptics can audit a report," Ries says.

This post was originally published on Smartplanet.com