Big (Bad) Data

By Mark Roberti

A lot of companies offer solutions for managing big data, but few admit that much of the information they manage is bad.

Thirteen years ago, I became interested in radio frequency technology while reporting on the problem of bad supply chain data, which caused faulty forecasts that led to problems. I was working for The Industry Standard at the time, and I would ask companies that deployed supply chain software from Manugistics and i2 Technologies whether they achieved the expected benefits. Invariably, the answer was "no."

When I asked why they hadn't achieved the hoped-for benefits, the response was always the same: It wasn't a problem with the software, but rather with the data. People would enter incorrect information, scan the wrong bar code or whatever. This led to bad data entering the software, and, well, you know the old saw about garbage in and garbage out.

I set out on a path to discover which new technologies might solve the "bad data" problem. As fate would have it, I was at a conference that had nothing to do with data or the Internet, and was talking to a manufacturing executive about my pursuit of a solution to bad data. He said, "You should look at RFID," explaining that the technology lets users capture data automatically, taking human error out of the equation.

I bring this up because as I walked the exhibit hall last month at the National Retail Federation's Big Show, in New York, I witnessed a number of companies offering "big-data" solutions. Many firms were offering not only to store the wealth of data retailers now collect, but also to analyze it. Wow! We'll look at information that studies show is wrong 35 percent of the time or more, and we're going to tell you... what exactly?

Many of the solutions displayed at NRF, including automated checkout, self-service kiosks and customer experience technologies, either require the kind of highly accurate data that radio frequency identification can provide, or would work much better if an RFID system were already in place. I'm told that executives from Macy's actually told a group of high-level retailers that omnichannel retailing cannot be accomplished without RFID, because a company can never have the inventory accuracy required without it.

Information is the lifeblood of most businesses. I understand that because it is critical to RFID Journal's business. But if your store employee looks at a computer and tells a customer that you have an item the patron wants, but you don't actually have it—or, if your inventory system thinks you have 10 of a particular product but you actually have none—then all the big-data analysis in the world won't help you much.

Maybe next year, the big fad at NRF will be clean data... but somehow I doubt it. And that's because the only way to get good, clean data about your inventory is to use a technology most exhibitors at NRF don't understand—namely, RFID.

Mark Roberti is the founder and editor of RFID Journal. If you would like to comment on this article, click on the link below. To read more of Mark's opinions, visit the RFID Journal Blog, the Editor's Note archive or RFID Connect.