Big data is hype and most advocates are not even statisticians or IT savvy

Billions of dollars are invested in unimaginable amounts of data and data infrastructure. But I believe most of this is hype and companies will not learn how to use it.

Analytics generated insights is really cheap. What is the point of collected terabytes of data when a simple sampling exercise can produce similar insights? If you want to learn Asian shopping behaviour, do you need sensors and smart detectors to generate tons of data per minute? Or could you achieve similar insights with a group of part time surveyors?

A company must know how to effect changes to create value. The ability to collect and analyse data is not a competition advantage. Most companies don’t even need new infrastructure because their firms are already producing so much data!

Most big data advocates are not even familiar with basic math and statistics. Before you become a big data advocate for your firm, make sure you know what specific outcomes your codes can product.

Big data is absolutely useless for companies which may not be interested in evidence based decision making. You can own the data in the world, but if your CEO believes he is a prophet, big data is totally useless.

There’s only one area that I believe data will add value easily – supply chain and cash flow cycle. One can easily collect all available data and find out the specific process that slows down the value chain. Managers can also find out which workers are performing the worst and which accounts typically pay the slowest. But this is not “Big data”. This is ability that we had since the beginning of days, or maybe, since Microsoft Office 1995.

To me, big data is all hype. The difference between data processes today and 20 years ago is the speed that we produce and analyse data, both enabled by machines and codes. In short, we are merely running the figures quicker than before. And like I said, most advocates of big data have not coders, mathematicians, statistician or technology engineers.