Orphaned data — information that companies collect and store with the intention of analyzing it at some future point — may have a name that tugs at your heartstrings but more likely than not, orphan data is obsolete and dragging on your organization’s purse strings instead.
That’s because orphaned data is particularly expensive to maintain according to a 2016 report from the Data Genomics Project. Orphaned data accounts for about 1.6 percent of the total enterprise file population but hogs about 5.1 percent of valuable enterprise storage capacity, all while eating up overhead and inflating IT operating costs.
What’s more, orphaned data is ownerless and stale: More than 40 percent of corporate data hasn’t been modified in at least three years and about 12 percent hasn’t been touched in the last seven years.
Perishable Insights Gone Wrong
Yet no organization sets out to create or hoard orphan data.
More often than not, what happens is that data is captured with the goal of delivering perishable insights from real-time analytics — think clicks on a website or geo locations from mobile customers —but that data is only useful while that customer is making a purchase.
If action and analysis do not come together in that single moment, that data has lost its value as real-time information. And weeks, months — even years can go by without an organization even realizing that its orphan data is there.
Catch and Release in the Data Lake
What’s an organization to do?
For starters, enterprise data needs to be classified and tagged, not simply squirreled away in storage to be quickly forgotten. Through classification and tagging, some orphan data can be repurposed as historical data and re-analyzed to derive management insights to improve business results.
Once classified and tagged, useless orphan data can be pulled permanently from the data lake while still-valuable data can be thrown back to play a more useful role in the organization’s big data ecosystem.
Taking the Fast Data Approach
However useful data classification and tagging efforts may be though, it’s much more efficient to derive value from data before it becomes orphaned and goes dark by implementing a fast data solution that ingests, analyzes and acts on live streams of data in real time.
This allows an enterprise to take action, automate transactions, and make decisions based on that live data stream. An enterprise can immediately analyze and act on real-time information regarding sales, production and distribution trends, as well as capitalize on opportunities to create value by shaping products and steering marketing and sales campaigns in real time.
Reducing Orphans, Increasing Agility
Maximizing the chances of capturing value from data before it goes dark requires adoption of an information management strategy that accommodates unstructured and structured data with consistent tagging and metadata policies.
As a general rule, any batch approach to storing, classifying, and analyzing data not only increases the odds of creating orphan data but reduces that data’s ability to contribute to agile, real-time decision making.
Fast data pipelines allow the enterprise to understand business impacts in milliseconds and automate actions based on policies to turn soon-to-be orphaned data into valuable enterprise information. By applying a fast data approach to incoming streams of data, enterprises can extract real-time information.
Enhancing Business Decision-Making
Instead of allowing low-value data to consume enterprise storage resources without driving improved business outcomes, the enterprise can ingest relevant fast data sources into a fast data infrastructure to enable real-time analytics.
Using that approach, organizations can more quickly and accurately assess customer requirements, detect fraud and assess risks in real time. The enterprise can gain immediate insights to shape management decisions that will nurture sales and drive business growth, and dramatically reduce not only the existence of orphan data but the financial and decision-making costs of storing it.