Nine Big Data Facts to Guide Your Technology Resolutions

By January 5, 2015Data Center, Storage
nine big data takeaways for 2015

We recently came across one of Gartner’s newest reports, “Major Myths About Big Data’s Impact on Information Infrastructure,” which explores nine big data myths. An abundance of marketplace hype and news about big data has caused a series of misconceptions surrounding the hot topic. To help straighten out the confusion, Gartner completed valuable research around nine major areas. We read through the lengthy report and compiled the biggest takeaways for you.

If you’d prefer to wade through the report on your own, check out the full Gartner research report instead.

Nine Big Data Takeaways for 2015

  1. Everyone has not already adopted big data. Many organizations exist in a state of fear that they will be the last to develop and implement big data strategies. Thanks to Gartner’s research, we know that in reality only 13% of organizations have actually fully adopted big data solutions. On the flip side, the research also reveals that close to 25% of organizations have no plans to invest in big data at all.
  2. Quantity doesn’t trump quality. Organizations are becoming so obsessed with collecting big data, they’ve forgotten about the importance of quality. Gartner explains that many organizations think that larger data sets mean flaws have less of an impact. In actuality, the number of flaws grows as the data does, meaning the overall impact of errors is unchanged. IT organizations need to put just as much effort, if not more, into big data quality strategy.
  3. Big data won’t eliminate data integration. According to Gartner, using new technologies to replace some data integration steps won’t eliminate the need for integration; it will just change the design. Gartner recommends assessing the existing architecture to pinpoint areas for improvement as top priorities for new engineering approaches.
  4. Data warehouses will still be relevant. While a lot of hype surrounding big data seems to suggest there’s no longer a need for data warehouses, Gartner reports that’s simply not true. According to Gartner, organizations don’t necessarily need to utilize data warehouses during the experimentation phase, but they should still implement them afterwards. Gartner points out that curated data leads to the best scoring and risk models, keeping data warehouses relevant.
  5. Data lakes will not replace data warehouses. Critics are also arguing that data lakes will replace data warehouses. Garter responds by suggesting that data lakes aren’t quite as easy as they seem. They recommend capitalizing on already-successful data warehouses, instead of dealing with the lack of analytical and data-manipulation skills necessary for data lakes.
  6. Hadoop will not replace data warehouses. With a third theory, critics suggest that Hadoop will replace data warehouses, but Gartner dispels this one too. Gartner reports that fewer than 5% or organizations are actually planning on replacing data centers with Hadoop, and the number is actually decreasing.
  7. Big data has not fully matured. As specified earlier, only 13% of organizations have actually deployed big data solutions. While big data is growing, it’s not a mature market and there is still an incredible wealth of information that is unknown, making risk very realizable.
  8. Big data is not the end-all, be-all. While big data seems big, it’s not the first opportunity for IT development. Gartner’s research shows that processing capacity doubles every 22 to 28 months, memory capacity doubles each year and every five years storage improves in efficiency, among other advancements. Instead of reacting in a panic, Garter suggests that organizations should calm down and develop a logical plan that works for them.
  9. You need to keep data governance in mind. Because big data is just like any other data, only larger, organizations still need to employ data governance initiatives. Gartner mentions that as organizations implement big data they should re-evaluate the current data governance structure and make sure it is flexible enough to meet the criteria needed.