How organizations can make cold data warm and valuable
Data is the new lifeblood of the economy and the fuel for innovation. With data insights poised to transform virtually every industry – from healthcare to manufacturing – it’s no secret that companies, entrepreneurs and visionaries alike are looking to data to advance their respective industries, with artificial intelligence and machine learning acting as the catalyst.
While industry practitioners remain optimistic, a large percentage of companies will need to rethink the way they view and store data within their organizations. Today, data is not easily accessible and thus the business lacks the ability to extract value. Most companies have stashed data in “virtual piggy banks” or data silos that have built up over the years.
My daughter Margot, age 4, likes putting all her change into her piggy bank. However, in the last few months, it dawned on her that once the change is in the piggy bank, it is hard to get it out and play with.
Her change is like today’s data. It is locked inside a piggy bank until she makes a hard decision, which entails spending time and energy getting the piggy bank open or breaking it altogether. Neither is an ideal option. However, she still wants to take that data – or her change – and put it into her application, otherwise known as her play cash machine.
My daughter, like most companies, has a hard choice to make to get the result she seeks. While my daughter contemplates her next move, one thing is certain for companies – a locked-in data approach creates more problems than solutions. In fact, this approach transforms data into a company’s most expensive asset rather than its most valuable asset.
While this analogy may paint a doomsday picture of today’s data pipeline, there are approaches companies can take to address data silo roadblocks. First, companies need to select a data storage model that will allow data to be easily extracted and analyzed. One avenue to consider is implementing a data hub into a company’s infrastructure. A data hub is a data-centric architecture that allows data to be freely shared and delivered to any infrastructure or application in real-time. The architecture was built on the idea that the data companies use for warehousing is also important for AI and other forms of analytics.
Second, companies need to analyze how they measure their data storage efficiency. A metric leveraged by industry professionals in recent years has been $1 per gigabyte of data. While this metric used to make sense when companies were simply storing their data for a rainy day, today’s need to extract and act on data rapidly has created a business environment where $1 per gigabyte is not a feasible metric to measure success.
Instead, companies should view data storage efficiency with an outcome-oriented lens rather than compute-oriented. This change allows the metric to become dollars per simulations. Moving to an increased focus on outcome orientation benefits companies because the real value of data is now generated by simplifying and accelerating the data flow instead of measuring complexities in the storage process.
While addressing data silos that have been architected throughout a company’s history may sound like a huge undertaking, the benefits in doing so are well worth the time and effort. The utilization of “warm” data has allowed industries to revolutionize traditional methods and deliver exceptional service to their end users.
Healthcare provides a great use case, where industry pioneers are utilizing AI in cancer pathology. Practitioners are now training AI technology to focus on digital images of cancer biopsy slides, some of which were collected over 30 years ago, to help doctors diagnose disease.
This is an exceptional time to be living and working with data. There has not been a comparable time in history where data that traditionally lived in a company’s IT department can now help oil well loggers receive detailed insights while measuring thousands of feet underground or help a car drive through city streets without a human regulating speed or steering.
To realize these truths, companies need to take a hard look at their data storage processes and ensure their data is actionable. The days of warehousing data for the sake of warehousing are long gone. Confirming data availability and actionability is the first step for companies to innovate their business and pave the way for increased industry innovation.