Big Data: Breaking Excel, and beyond!
Mainframe is one of the most overlooked Big Data tools out there. But there’s a lot more that needs debating. Join our panel chat on June 27
The third installment of CA Technologies Google Hangout on Air panel series focused on Big Data, and included a number of great points that I would like to spotlight for further discussion. We have already covered a lot of Big Data-related topics, so I’d love to bring that together under the umbrella of a litmus test of the magnitude of Big Data – “Could it break Excel?” – and how that ties back to the success and usability of Big Data in business.
Now, we probably could have done a three-part series on defining Big Data, something our panelists strived to do over the course of the discussion, but one of the definitions shared really stood out to me. Panelist John Lucker, Global Advanced Analytics & Modeling Market Leader, Deloitte Consulting, defined Big Data as “optimally combin[ing] internal and external multi-structural data from multiple sources.” So much of that mission critical data resides on the IBM System z, and it plays a crucial role in all things data to scale, which takes me back to the “breaking Excel” concept.
In many companies, Excel is used to combine and present data from many sources. For example, it could be used to extract content from databases with different platforms, and then combine that data set with data from on-premise applications in addition to cloud solutions. These already “insane” quantities of data can then be simplified to subtotals, totals and averages.
As such, we all have experienced how this oversimplified data can be misconstrued when poured into pivot-tables and then transformed into PowerPoint presentations with easy-to-understand graphs and tables – the abridged conclusions often don’t ladder back to the raw data. Now imagine what would happen if we layer in massive amounts of external multi-structural data such as social media, weather, geographical, etc. If an analysis of that data gets just one average total wrong, every result drawn from that analysis is questionable at best.
So if Excel isn’t the best tool to amass and analyze all of this BIG data – then what is? One point all of the panelists could agree on is that the mainframe IS a source of very reliable data. The fact that so many businesses overlook the mainframe as this critical tool is more of an image problem than a functional problem – it’s one of the most overlooked Big Data tools out there. A number of companies do see the Big Data opportunities in their mainframes though, and have hired data scientist to unlock that data potential. As these data scientists explore the analysis capabilities of the all-powerful mainframe in an effort to optimize an overarching data strategy for the company, that data-ologist may run into hurdles in communicating this data potential to the more business-minded, less data-focused C-suite executives.
Data scientists and the C-suite must find a common language when it comes to Big Data, and more so than pivot-tables and simplified graphs/tables. The best way to cross the data barrier with the business-minded executives is to use that multi-structural data to demonstrate a REAL, TANGIBLE business value. With that, I’d like to make an addition to the Big Data definition I shared earlier… optimally combin[ing] internal and external multi-structural data from multiple source TO DELIVER PROVEN BUSINESS VALUE. What is Big Data to a business if it doesn’t bring value? Not anything to write home about…or even a blog post about!
There were a few other really critical points this panel touched on that I would like to call out:
Finally, I would like to add one last point – make sure you hire the right people to ensure you do the right type of analysis before you start making important business decisions centered on these data-based insights. If these projects go well, Big Data can evolve from hype to an invaluable business tool. So, check out the full panel below for some great Big Data discussions, and until next time, to Big Data, and beyond!
Please join our next panel in the series DevOps: Understanding and Investing in the Sweetspot of IT June 27th at 11 a.m. EST