“Would you rather have a surgeon who knows everything from books or surgeon who has actually done your kind of operation?” That is how Andrey Aredakov sums up his teaching philosophy in the master’s degree in business analytics program at GGU. “Knowing a formula for standard deviation doesn’t help much in business. One has to apply it to solve concrete business problems.” For this reason, Aredakov gives his students a data set and challenges them to unlock its potential and communicate the results. In the real-world of business, these insights will give their colleagues what they need to make decisions.
Aredakov works as a Data Analyst at a major search engine company based in the San Francisco Bay Area. Colleagues from various sectors of the company approach him to make sense of an enormous data flow that is expanding by the millisecond. This data flow emerges from user behavior, digital advertising customers, and information from mobile devices. Aredakov presents data results to the various “clients” in the company such as product managers, project managers, and engineers—as well as members of the finance, marketing, and sales teams.
“Storytelling and good communication are part of the job of a business analytics or data analyst professional,” says Aredakov. “This is not about the book knowledge or the statistics. It is about taking out what is unnecessary and telling a kind of story to colleagues that will help them grow the business. The trick is not to create many slides that are full of data charts, but come up with one slide that leaves out everything extraneous and makes finding the solution easier.”
Transforming data into a story liberates a fundamental business asset and gives professionals what they need to improve a company’s product, efficiency, and bottom line. “In today’s world, we have a lot of data and computers that help us compute anything in a short time,” Aredakov says, “but all these numbers require interpretation. Only an expert can provide that.”
Aredakov’s students are using R, an open-source statistical tool for business that is dramatically simpler than query languages such as SQL and scripting languages such as Python. You can get results sets, such as the “correlogram” below, in a matter of seconds just by calling the function. The example below correlates car MPG performance (top left box) with various factors such as the number of cylinders (column 2).
About Andrey Aredakov
Andrey Aredakov has worked as a quantity/quality researcher in media, web, and advertising analytics for the last twelve years. His specialties are media and Web analytics, as well as competitive research. Aredakov earned a PhD degree in philosophy at Moscow State University. After working as a data analyst at various companies in Toronto, he moved to San Francisco and joined a major search engine company.