- For Financial Services
- Dev Center
- About NuoDB
- Get Community Edition
- Watch the Demo ❯
You are here
Jump Aboard the Data-Driven Train
Mar 2 2015
The Data Driven train has already left the station. That is, at least, according to the recently announced preliminary results of the EMA/9sight 2014/15 Big Data survey. Over 60% of the respondents claimed that a data-driven strategy was already adopted in their organizations, and a whisker under 20% recognized it as a vital part of their business. So, if you haven’t boarded the Data Driven train yet, it would seem that you are in danger of being left behind.
But, what does it mean to be Data Driven? Haven’t we been using data to run and track our businesses since time immemorial? And I mean long before Business Intelligence became the buzzword of the 1990s, never mind the recent Big Data brouhaha. Being Data Driven does NOT just mean having access to your customers’ Facebook or Twitter comments. Nor does it just mean tracking their whereabouts via their smartphones and pushing location-specific advertising to them. Of course, it may include such modern manifestations and more, but that’s not what it’s really about.To be Data Driven is to exhibit three characteristic behaviors around the management and use of data within the organization.
First, you must have a clear and well-implemented strategy around what data you want to collect and keep. Today, the ground rules have changed from “what can we afford to keep?” to “what can we afford to miss?” That’s not to say you should collect and keep everything, as the Data Lake mindset might suggest. Depending on the value and usefulness of the data, you should exercise the full range of options from “do not collect”, through “collect and dispose”, to “collect, preserve and manage carefully”. The challenge, of course, is to predict which data may have future use and value, even though it may have none now.
Second, a data-driven organization must have a comprehensive data architecture that is capable of handling a wide variety of data types and potential uses. This includes both relational and non-relational data, highly or more loosely structured data. It must cater for operational, informational and mixed workloads, running on in-memory databases, Hadoop file systems and everything between. And, with such a distributed storage architecture, special attention is needed to maintain consistency and cleanliness across the multiple stores.
Third, the widespread use of data across the whole organization in all aspects of the business must be enabled and encouraged. This means having the right tools and the right data for the job in hand. Appropriate levels of data governance and management must be applied to drive high confidence in data-driven decisions. It does not mean the elimination of judgment or intuition where appropriate, but it does insist that such judgment or intuition is applied only after the data has been investigated.
The bottom line is that being Data Driven is a strategy that spans the entire lifecycle of data within the organization and demands attention from both IT and business users alike. Now, go catch that train!