Data scientists and analytics experts are much in demand and hard to locate, but how can you be sure you've got the best candidates, and what should you ask them? Google — obviously no slouch when it comes to analytics — has some useful suggestions on the best approach.
Analytics picture from Shutterstock
Google's "analytics advocate" Adam Singer spoke on the topic of how to successfully locate and hire analytics experts during the MozCon 2015 conference in Seattle, which I'm attending this week. Singer's biggest theme was a familiar one: you definitely need analytics expertise, but if you can't find (or afford) someone who is a full-blown data scientist, other options will do.
"The best analysts are comfortable working with imperfect data," he said, noting that online data is often incomplete and fragmentary. "Having a stats background actually can give you a handicap. You don't need to be a scientist to be a really good analyst."
Knowing your packages is helpful, but the human element of analytics is still vital. "Software is not at the point where it can do your analysis for you," Singer said.
That said, technological knowledge is still important. "Analysts need business acumen and nerd skills," Singer said. That's a reassuring thought, given how often we're blithely told that IT pros need more business expertise rather than more product knowledge. Actually, you need both. (You also need a relatively high tolerance for danger — working in science-related fields can be riskier than you might think!)
Assuming you assemble a candidate pool that meets these requirements, Singer suggested three critical questions to ask during job interviews:
We have several (varying) groups that need access to analytics insights, from product to marketing/PR and customer service. How would you efficiently get them all what they need? This question ensures that your prospective hire has a sense of how they would balance the incessant demand for analytics, and can also give you a sense of their interpersonal skills.
How do you see a breakdown of time spent on analytics between data capture, reporting and analysis? The breakdown will obviously vary — a business that already has capture systems in place should be able to spend more time on analysis.
How do you plan to be data-driven in improving your digital marketing efforts? Singer proposed the question in this form because MozCon has a heavy emphasis on technology for marketers, but the same formulation could easily be changed to reflect your business focus: "How do you plan to be data-driven in expanding our sales/speeding our development pipeline/increasing our audience?"