As the “big data” juggernaut* continues to build momentum, it is about time to pause and reflect on those things that are holding us back from realizing the true potential of data integration, overlays, and analysis. Mark Miller posted a great piece on MediaPost (“Three Steps To Dealing With Data Paralysis,” Jan 31, 2012) that points out a few of those issues and how to circumvent them.
- Favor smart data over big data: “Data issues like missing values, missing linkages and data anomalies can impede your ability to harvest data and move the business forward…. have a plan to deal with quality issues that will otherwise make your data dumb.” Hallelujah.
- Use analytics to mine your smart data: “…identify the 10-15 most important things you need to know… Focus on the information that helps you determine…how you can create value.” Setting the right data priorities is, of course, key.
- Create a roadmap using data and analytics: “…conduct an audit of the processes, systems, tools, and talent within your organization. Identify the gaps. [This audit] should be connected to harder metrics like sales, revenue and profit.” There a couple of good points here: 1) If you don’t have the talent or bandwidth in-house for the data work then go get it on the open market; and, 2) Always focus on “real” numbers — confusing “reporting” with “analysis” is a fatal mistake and focusing on hard metrics (not forgetting that cost reduction is among those) is essential.
Miller comes from the CRM world, but what he is saying is relevant to many others in the world of creating and managing data. We may all be awash in data and have decades of experience under our belt managing it, but it’s never to late to go back to basics and make sure that the foundation for our databases is well laid, our cost of ownership is low, and we are getting every dime of value out of what we have built.
*Derived from various Indian words for a very large, portable, wooden temple.