Guide to big data analytics tools, trends and best practices
A comprehensive collection of articles, videos and more, hand-picked by our editors
Every business that accepts cash payments is susceptible to the problem of shrinkage: when employees or customers intentionally or unintentionally misplace money or merchandise. But some businesses have begun mining big data to fight the problem.
Mississauga, Ontario-based Compass Group Canada, an owner and operator of more than 2,000 food service locations, like Tim Hortons, Subway and Starbucks north of the border, recently implemented a self-service analytics engine to help make correlations between disparate data systems and put an end to the problem of shrinkage. Company officials said this approach is an improvement over traditional efforts to stop theft, which mainly rely on watching hours of video footage.
"There's millions and millions of hours of video that gets generated on a monthly basis," said vice president of business transformation Humza Teherany. "To go through it, you just couldn't manage that. We actually catch people on data first and then correlate it to video."
The organization implemented software from Boston-based Lavastorm Analytics. The analytics engine works on data from a common database that collects data from Compass Group's point of sale systems, enterprise resource planning system, employee time and attendance software and inventory management systems. The analytics software combs through this data for correlations between register discrepancies and other information. For example, if a drawer is short $20 every night after one cashier works, the loss prevention team will know to watch the security camera footage from that person's shift.
This type of correlation may seem obvious, but Teherany said with over 2,000 locations and even more employees, it's like finding a needle in a haystack. Prior to implementing the analytics engine, Compass Group employed teams of internal auditors. But these teams had to manually go over endless spreadsheets. By the time they spotted anything irregular, it may have been too late.
However, since Compass Group started mining big data, it has been easier to identify employees who were potentially involved in theft, Teherany said. In the amount of time the software has been in place, Compass Group has opened 10 times as many case files as it did previously over a comparable time period prior to implementing the analytics tool. Furthermore, 100% of cases identified for further investigation by the software have been confirmed by other evidence as instances of employee theft.
We actually catch people on data first and then correlate it to video.
vice president of business transformation, Compass Group Canada
Teherany said this success has boosted the confidence of the company's internal audit team, the primary users. They are now feeling more confident about their ability to be productive at stopping shrinkage.
"That's what the analytics approach has allowed us to do, is reenergize people and show people that we are going to get on top of it," he said. "I'm not naïve enough to think that we are going to solve the issue, but just reducing it would be huge for us."
Mining big data for loss prevention purposes is likely to become an industry standard in the near future, said Brent Mooney, chief financial officer at Compass Group. Particularly for large organizations like his, there is simply too little oversight of day-to-day operations, he said. Even security cameras rarely act as a deterrent, because people generally know that no one is actually reviewing the footage on a regular basis.
Due to the amount of money that is at stake, Compass Group did not view the issue as a human resource problem, Mooney said. Also, they didn't view the selection and implementation of the analytics engine as an IT issue. Instead, they looked at it as something the entire business was responsible for solving. To that end, they formed a loss-prevention team made up of people from HR and legal, which was headed by Teherany. Mooney said this group was tasked with finding ways to slow shrinkage. It eventually determined that the company could use its data to stem the problem and selected the software.
"We looked at all of this as a business issue, not as an IT issue," said Mooney.