Calculating Crime Risk
Cooperative food retailer digs deep into data before investing in resources
By integrating proprietary operational and crime-related data into a crime risk data model, U.K. group Co-op Food has been able to significantly reduce crime, theft and operating losses.
Owned by more than 8 million members, Co-op Food is the United Kingdom’s fifth-largest food retailer, operating some 2,800 local, convenience and medium-sized stores across the country that serve more than 14 million customers each week. The stores range in size from 1,000 to 20,000 square feet, and are in both rural and urban locales.
The variety of stores and environments presented a challenge to the organization in deciding how to allocate resources. “We were finding it increasingly difficult to apply an objective view of risk and to prioritize our investments to best allocate the limited spend available to us,” says Graham Watt, improvements and projects manager for the Co-op.
‘A clear view’
Co-op Food began implementing the risk model, developed by consulting firm CAP Index, by using the data it was analyzing primarily to help allocate security resources to the stores that needed them the most.
It first integrated historical operational and financial data with CAP Index’s risk model, which combines demographic statistics and actual crime data to measure an area’s social disorganization.
That addressed the Co-op’s need for “a more organized and consistent manner of evaluating overall risk at specific locations and applying their resources accordingly,” says Stephen Longo, vice president of strategic initiatives for CAP Index,
Over the years that Co-op Food and CAP Index have been working together, the Co-op has been feeding CAP Index “a lot more data sets,” Watts says. “That allows us to create an operational model as well as a crime risk model. When we blend the two, they give us a very accurate overview for each location.”
The Co-op now has more than 200 datasets, including store-level criminal incidents, turnover, current security expenses, unrecorded losses, labor costs and security investments to “give us a clear view of what is happening in each location,” Watt says.
That is aligned with CAP Index’s objective crime risk data, and then statistical modeling is used to create an accurate business crime risk for each location.
Today the Co-op is feeding in data that is helping it identify and prioritize risks that arise in many financial and operational areas.
For example, how much does it cost the Co-op when it suffers stock shortages because of theft or inefficient processes? Or if the appropriate level of security at each store is not maintained?
The organization is even addressing the challenge of how to best control risk-related investment expenses and insurance costs. Managing all those business needs “complements the CAP Index risk model very well,” Watt says.
The data allows the Co-op “to scope out stores impacted by external factors and stores impacted by internal process issues, which allows us to forecast our activities in the right areas.”
The Co-op uses the model to support what Watt calls an “aggressive planned acquisition program, enabling us to establish the cost to operate stores in given areas,” allowing management to determine whether to proceed with an acquisition or “walk away from it.”
It was due to the objective analysis of the CAP Index crime risk model used to evaluate the investment in next-generation security cameras that the Co-op decided to test the technology in 160 stores.
“In a world of limited resources,” Watt says, “typically one of the first things to be removed from retrofits … is security equipment. The custom risk model allows us to objectively determine when, where and whether we should invest more rather than less.”
At the end of 2015, he says, violent crime in the elevated-risk stores allocated with the new equipment had been reduced by 60 percent. All crime reported was reduced by 40 percent and shrink performance had improved by 19.5 percent.
“As a tool for understanding where we needed to invest and in deciding the package we should invest in,” says Watt, “that alone made a very strong business case for using the crime risk model. It helped channel the money into the right places.”
Not only has the reduction in losses resulted in a “substantial return on investment,” says Watt, it has also simultaneously “impacted customers in a positive way.”
In stores with reduced crime risks, Watt says employee satisfaction has increased by 7-10 percent, and customer satisfaction is up modestly over prior-year levels. He believes that’s partly attributable to the fact that the risk model has enhanced in-stocks, which means product availability has improved.
The Co-op recently began identifying where it needed to invest in employee behavioral training. “We want to deter crime rather than react to it,” Watt says; the company has begun allocating more manpower to elevated-risk stores.
The company has also begun shaving costs by using the model to determine how frequently individual stores need to be inventoried. Some stores counted stock twice a year, others three times. By using the model to identify stores where stock could be counted just once annually, inventorying costs were reduced by as much as 50 percent in many instances, Watt says — saving more than $1 million.
The Co-op is now looking to reduce its insurance premiums based on its ability to cut waste and criminal incidents following its targeted investment activities.
Over the last five years, the Co-op has worked hard to make sure the data it puts into the model is “clean and accurate,” Watt says, “and that the data sets that have been added allow us to do far more than just guarding.”
The organization previously applied budgets based on previous performance, “so we almost rewarded poor performance on a year-to-year basis,” he says. “This allows us to objectively look at any area and identify elevated-risk stores and budget accordingly. We can identify stores performing poorly in low crime-risk areas and identify those having poor processes. That allows us to invest in improving processes to improve performance.”
Watt says some of the programs have returned the investment in less than four months.
“Being able to target 2,000-plus stores and save some money in all of them is far more beneficial to bottom-line performance than focusing on a couple hundred stores and trying to get big money out of them in a really challenging environment.”
Moving forward, he says the Co-op will “continue to refine the model. Our business is changing and the U.K. retail proposition is changing, and as our whole customer journey changes, the challenges we are going to have will change with it. So we need to be forecasting what is going to be coming.”
NRF members come from more than 45 countries and all sectors of retail, from Main Street merchants to online retailers.