You’ve probably heard about big data being used to figure out what you like to buy, read, and follow. What you likely haven’t thought about is how your company might use it to unleash your productivity.
But Alexander Vorobiev, an Advisor of Advanced Analytics at TransUnion, has. He’s a whiz at all things big data. And while his role primarily deals with how big data can impact financial services, he knows the applications for big data are endless. One such thing? Figuring out how companies can use analytical methods to increase productivity, and see better business results.
Sound intriguing? Read on to learn how it’s done:
Find Your Hypothesis
First you need a theory to test. “Creating a workplace wellness program will increase productivity” might be one. “Allowing employees to work from home will help drive sales” could be another.
As the head of a department or decision maker, you might have a gut instinct about how your employees work best. Maybe it’s that employees who come in one hour later take less breaks throughout the day, or if employees use their lunch hour to exercise they tend not to give in to the 3 PM slump. Whatever the assumption may be, this is your hypothesis to be tested.
Gather the Right Data
Arguably one of the most critical steps in using big data. All the analysis in the world won’t be of much use if you don’t measure the right things. Take the hypothesis “working from home improves productivity.” A few potential data points to measure here might include number of telecommuting employees, how many days they worked from home, and supervisor reviews at the end of the estimated period.
Vorobiev recommends that companies hire specialized data engineers or external consultants to conduct analysis of workplace trends and other areas where big data is sure to be of use. Such data scientists can not only analyze the final results, they can also suggest the correct parameters to measure.
Set Up a Sample to Study
Companies can recruit employees for studies by dangling a carrot (free gym membership for a year is a good one) although one has to watch out for biased samples (people who sign up for a book club, for example, might already be ones who like to read).
But recruitment can take place in other ways. Vorobiev points to a workplace study conducted by Bank of America where employees wore ID badges with RFID tags and their interactions with each other and subsequent productivity was measured.
However, Vorobiev admits that privacy is a legitimate barrier. But there are ways that hide employee information so analysts only focus on larger trends. Anonymous bubble answers or online surveys are a quick and easy way of looking for patterns without naming names.
Once you’ve figured out whom to study, online surveys are a quick way of gathering the needed data.
Now that you’ve got the results, big data can analyze it and look for trends. It is important to remember that big data analysis is simply regular data study on steroids. You, as an employee or company owner, could always conduct data analysis. But big data processes information coming from a number of sources and many different ways more efficiently and quickly.
Just don’t get lost in analysis paralysis. “You can over-engineer anything,” Vorobiev says, “There’s a famous saying about statistics that with enough pressure the data will admit to everything. It might be a good idea to stop analysis as soon as you get x numbers of input or results and then see what the data tells you.”
The streetlight effect— where a person who has lost his keys looks only under the light because that’s where it’s easiest to do so—is a legitimate concern when it comes to big data analysis. Remember that the most surprising trends might not be where you first think to look.
The takeaway according to Vorobiev: “There are so many measurable, easily overlooked, aspects of our work lives that, if studied, might produce unexpected results. And if one of them could lead to a more harmonious and productive environment, it’s worth trying.”