Companies face the challenge of making sense of vast amounts of data.Getty Images/iStockphoto
We explore 10 key challenges for business leaders in 2014, with expert commentary on the issues. For more stories on managing: tgam.ca/managing
When Jean-Paul Isson started at Monster, the popular job recruiting website, he was amazed to see just how much data the company was collecting. Every résumé, e-mail and Web page visited, among other things, offered invaluable insight into how job seekers used the company's services. That information, if analyzed right, could help recruiters create better postings and attract the right talent.
There was one problem: It was impossible to organize all of that information into something Monster could use. "Data integration was the real challenge," says Mr. Isson, Monster Worldwide Inc.'s global vice-president of predictive analytics and business intelligence.
He remembers asking four different people how many customers the company had after he started there nearly a decade ago, and he got four different answers. They weren't necessarily wrong, he says, but there were so many ways to interpret data that there were no unified answers to his questions.
More often than not, people looked at a few data points and then acted on a gut feeling, he says. They'd see some information and interpret it based on what they thought was right. The company also took a "sampling" approach to analytics – one that looks at a small set of customer data to determine what everyone else is doing.
While sampling did help the company understand its data and customers better, it doesn't take into account any anomalies. "There may be a reason why a job is not performing well, but if the reason is specific to that one company, you'd miss that because you're only looking at a small sample size," he says.
If the company wanted to truly understand its customers, it would have to find a way to look at every data point it was gathering. "We had to standardize," says Mr. Isson.
That's easier said than done. Monster collects information from its many operations around the world, but it also gathers economic data, such as unemployment rates, GDP growth, turnover rates and more, in order to get an accurate picture of the global job market.
There are also a number of ways Monster could use its data. One of its business goals that it hoped big data would help with was targeting companies who would benefit from its service. It can't approach the millions of businesses around the globe, so it has to be choosy on whom it sends its sales staff to talk to.
By using all of that external economic data, plus information from its own site on what jobs people are looking for, Mr. Isson hoped he could determine the types of people seeking employment and where potential employees were looking for jobs. He'd then be able to narrow down the types of businesses that would benefit most from Monster's recruitment services.
Mr. Isson started looking into ways Monster could analyze all of its data in 2008. He knew then what a lot of companies are figuring out now: that information is critical to his company's success. If he could understand what his clients wanted, then he'd be able to better target their needs.
He ended up hiring SAS Institute Inc., a Cary, North Carolina-based analytics company to help. He worked with the company to help figure out what to analyze and which data points would be best to look at. The company's software also allows him to see results quickly and in easily digested graphs and charts.
Big data is becoming an even more important issue, though, he says. In the past, companies were trying to find ways to use internal data and information gleaned from Web traffic more efficiently. Going forward, it's "unstructured data," such as e-mails, tweets and Facebook posts that could elicit the best information.
"That's where we're trying to improve today," says Mr. Isson. "We want to be tracking everything in real time."
The experts weigh in
Here's what two experts say about Monster's strategies, and why this is a key issue for 2014.
David Soberman
Canadian national chair in strategic marketing at the University of Toronto's Rotman School of Management
Understanding the data all comes down to articulating what it is you're trying to do. Monster has a more complex problem because it's collecting information on employers and employees. What they needed to do was to think about who's interested in buying their product. The essential question for any company doing this is, "What customers do I want to serve and why?" Answer that and you'll know which data points to consider.
In terms of technology, SAP and Oracle are the companies who have had the most experience dealing with those issues. They can help process all of that data.
It's important that they keep improving these processes because of competitive pressures. Nowadays, other companies are doing this and doing it better. You'll lose market share if you don't do it right. It's also just increasingly becoming a cost of doing business. You can't operate a business any more without thinking about how you're going to manage and use that information to better serve your customers.
Ben Lorica
Founder of San Francisco-based Verisi Data Studio, a consultancy that helps companies with their big data needs
They're right to focus on unstructured comments. That's where we're seeing a lot of growth in the big-data industry. It's difficult, though. You have loads and loads of text and you have to figure out the key phrases. In many cases, you don't know what you're looking for. One way to start is by spitting out all of the possible phrases and counting them up to see what people are saying.
Look at companies like DataMeer, Platfora and Clear Story Data to help make sense of all of this data. They offer nice patterns and charts that make it easy to interact with the information.
Follow us @GlobeSmallBiz and on Pinterest
Join our Small Business LinkedIn group
Add us to your circles
Sign up for our weekly newsletter