We all love Big Data. Marketers eat it up when they see that something has been liked so many times on Facebook or that people are engaging in a Twitter hashtag.
But as Philip Wisoff, a CIO with Proskauer Rose law firm, said on our show “Addressing the Big Data Veracity Challenge,” “Big Data is not great at predicting human behavior,” and the amount of clicks and raw numbers that can be amassed by an algorithm can be misleading when we trust it without fail to govern our business solutions.
Big Data is as susceptible to flaws as old forms of BI, as Ram Akella, a professor on IS at the University of California Santa Cruz, said on Wednesday’s show, and it may be an even greater issue since we’re now gathering information at a quicker pace. There is however a difference between Big Data and “Data Sparsity,” as Akella called it. What’s missing is the promised action from the hits. We can estimate how many customers have seen our ad, but how many is it really affecting?
Right now, advertising on Facebook and other forms of social media is a bad barometer for such information, and it requires CIOs to really make sure we’re asking the right question when we’re using complex algorithms to gather data. Using basic human intuition can help eliminate all the data that is either incorrect, irrelevant or both.
As we wait for algorithms to get more reliable in predicting human action and approaching our business solutions, we can focus on what Akella calls “Extractive Analytics.”
“Given a ton of Big Data, how do you extract what is meaningful before further processing to get insights and action,” Akella asks. One day soon we will reach the point where we can reduce the cost and manpower required to analyze data without sacrificing the ability to make informed, helpful and human decisions.
Wisoff made a strong analogy in regards to trusting Big Data wholeheartedly. If presented with a Lamborghini in 1910, you might question whether or not what’s under the hood could really get you to 100 mph in three seconds. In present day, Lamborghini is a tried and true institution, and you can be assured that when you put your foot on the gas, you’re going to get results.
Big Data is still in its infancy. We’re still working out how we can process what we see on social media into efficient sales with our current clients, and we’re understanding how tinkering with SEO and Google ad traffic can boost our exposure with new ones. It’s not fool proof yet, but it’s close.
Listen to the full radio program “Addressing the Big Data Veracity Challenge”