Advances in data-analytics technology over the past decade offer enterprises new opportunities to mine critical business intelligence in near real-time, giving them valuable insights into customer preferences and behavior and uncovering previously undetected relationships and trends.
But technology alone isn’t enough for an effective data-analytics operation. Enterprises need skilled data scientists who know how to parse information and help draw conclusions that will benefit an organization. And that’s an enormous challenge.
“It’s difficult to find people who have a data-mining background and sufficient quantitative grounding in order to make sense of large data sets,” says Jason Hoffman, senior director of monetization for paid search advertising at Microsoft. “You have to do a lot of interviewing. We’ll have to spend lots of time going to conferences and get hundreds of thousands of resumes, conduct maybe a hundred or more phone screens, then we might bring in a dozen or so people and maybe hire one. And that process can take six or eight months.”
Microsoft’s experience hardly is unique. Truth is, the big data revolution is being held back by a lack of boots on the ground. The U.S. could face a shortage of more than 140,000 people with deep analytical skills by 2018, notes an oft-cited McKinsey report published in 2011. The report adds the country could lack another “1.5 million managers and analysts with the know-how to use the analysis of big data to make effective decisions.”
Enter Mu Sigma, a Chicago-based company that is among a list of consultants including Deloitte and Optaumum offering decision sciences and analytics services to companies with plenty of big-data needs but few or no employees capable of turning that data into actionable knowledge.
Started in 2004 by former Booz Allen Hamilton strategy consultant Dhiraj Rajaram, Mu Sigma aims for the sweet spot where applied math, technology and business interests converge. (The firm’s name is derived from the statistical terms “Mu (μ)” and “Sigma (σ)” which symbolize the mean and the standard deviation respectively of a probability distribution.)
“As data is doubling every 18 months, and the cost of memory is becoming cheaper and cheaper, the amount of signals that we all are receiving on a constant basis about each of us, about the environment in which we live, about the products that we use, about our friends, about our acquaintances, the signals are just exploding,” says Rajaram. “It’s not enough if you make decisions based on what you know. It’s important for you to make decisions based on what you can learn.”
Talent Pool in Bangalore
Mu Sigma has about 1,500 employees worldwide, with about 200 in the U.S. and the bulk based in Bangalore, India. The company specializes in three areas—marketing, supply chain and risk analytics. Mu Sigma boasts a client list that features more than 75 Fortune 500 clients in multiple industries, including Pfizer, Dell and, of course, Microsoft.
Hoffman says his group is one of several within Microsoft, which has been using Mu Sigma for more than seven years.
Hoffman describes paid-search advertising as a “very data-driven business” whose goal is to “improve the experience” for users, advertisers and web publishers.
“One piece of this is sort of a consumer psychology or behavioral economics problem,” he says. “Another piece of this is econometric, there’s an optimization aspect to it. There are machine learning and computer science problems. There are user experience problems to be solved. And the thing about this is all of those require the use of the data and qualitative techniques to get at the meaning behind those data.”
Not a simple task in the modern Internet era, Hoffman explains.
“There are other channels that people are using online besides the ones that were around four or five or six years ago,” he says. “There are different devices, there are different operating systems, there are different modes of interaction. All of this together becomes difficult to sort out.”
But sorting all of it out is imperative for a search operation that stakes its success on effectively gathering and analyzing data relevant to users, advertisers and content providers.
“For example, a traffic spike due to a news event, how do you sort that out? That’s not going to be the same as inventory forecasting for coolers on Fourth of July at Wal-Mart, it’s going to be a lot more complicated,” Hoffman says.
Beyond Mu Sigma’s ability to help Microsoft scale its data-analytics operations to meet the challenge of Redmond’s paid-search business, Hoffman says the analytics-services company is more client-centric than most hired consultants.
“Mu Sigma is different than most consulting in that they’re there to make the project successful and the project owner successful,” he says. “Often on the services side, they’re looking to move up the food chain and make the strategic sale, and often end up competing with internal teams. That’s not the case with Mu Sigma.”
Charles King, principal analyst for Pund-IT says “Mu Sigma’s analytics services offer a couple of attractive propositions.”
“First and foremost, the company helps customers get going with analytics without a sizable upfront investment,” King says. “That’s particularly important for companies with limited means and for those uncertain about the long-term value of and payback from deploying standalone analytics solutions. In addition, even if companies have decided to invest in analytics, the current shortage of data scientists makes it difficult or even impossible for many projects to proceed.”