That big data and analytics can transform enterprises in the digital economy no longer is a matter of debate or conjecture. Examples of big data success stories abound.
Yet it’s important to remember that big data and analytics aren’t magic, even if they may seem magical to enterprises successfully deploying them to improve their businesses. There are limits to big data that must be understood to effectively develop an analytics initiative and make sound business decisions based on results.
Over at Forbes, AppNexus chief data scientist Catherine Williams discusses (with contributor and AppNexus CEO Brian O’Kelley) why it’s crucial for enterprises not to have blind faith in big data and analytics alone. Another ingredient is essential. Or as O’Kelley puts it, “Without human intelligence, all the data and machine learning in the world won’t set you free.”
Williams agrees. “Bad things can happen when big data entirely displaces human judgment and discernment,” she says.
I would add that bad things also can happen when big data is combined with poor human judgment and lack of discernment. The truth is that the right analytics tools matter, but what matters more is the ability of humans to:
- Ask the right questions.
- Look at the right data.
- Interpret the data accurately.
That’s why, in my opinion, it’s critical to have an experienced data scientist overseeing an analytics program to ensure lines-of-business leaders aren’t misinterpreting data or asking the wrong questions.
It takes an experienced data scientist or analyst to separate signal from noise. Hard-charging and well-meaning business executives are apt to latch onto gaudy outlier data and draw inaccurate conclusions. Disastrous business decisions typically follow.
“We should be grateful to big data for its power to accomplish so many different things,” Williams says. “But again, the data should work in service of human beings, not the other way around.”