In the bimonthly Evidence-Based LP column, we’ve often discussed precision problem-solving. Precision means better outcomes with fewer negative side effects. And greater precision comes from a better problem diagnosis, meaning a more complete description of the very specific problem and its likely causes and where and when it’s clustered.
Recognizing one size does not fit all when it comes to retail risk management, retailers increasingly assign risk and vulnerability scores to their locations.
Risk estimates how much relative exposure a given store has to nearby clusters of surrounding offenders (the more likely offenders, the higher the risk) and how accessible the location is to these offenders. Store risk obviously varies widely, even within markets. Retailers subscribe to services that estimate area risk using reported crime and estimated social disorganization as examples.
Relative vulnerability is how well a store, distribution center, or office is capable of handling crime attempts. Every store’s ability to prevent and handle problems also varies. A place manager’s loss control knowledge and commitment varies. So does their asset protection toolkit. Historic loss, shrinkage, manager performance, reported incidents, and other metrics help retailers prioritize support.
The currently described retail risk management research was an earlier attempt to gauge whether and how retail chains segregate stores into risk and vulnerability bands for more precise protective support.
Study Method: Loss prevention executives from 21 companies in six categories (mass merchants, department stores, drug stores, apparel stores, specialty stores, and grocery and dollar stores) completed surveys.
Results: Almost all the retailers we talked to use some process and data to evaluate each store’s relative risk and vulnerability. Following are some study highlights.
By far, the type of risk data most likely to be collected by the participating companies is actual loss/shrink, collected by almost all (95.2%) of the participants. More than two-thirds (71.4%) of the loss prevention executives indicate they collect data on the number of incidents by crime, and a similar percentage (66.7%) collects data on the number of accidents at the store level.
Approximately four-fifths (81.0%) of the LP executives surveyed indicate they use the risk data they collect at the store level to make a risk profile for each store or to classify stores into categories. Nearly three-fifths (58.8%) of the companies who report using store-level risk data to classify their stores have three classification levels based on this data. Nearly one-quarter (23.5%) of these respondents have five classification levels based on store-level risk data. The most typical classification schemes are ordered number or letter categories such as “1, 2, 3, 4, 5” or “A, B, C” or categorical ranking schemes such as “low, medium, high.”
More than two-fifths (41.2%) of the participants that classify their stores based on store-level risk data determine which stores belong in each category based on a combination score of actual loss/shrink, number of incidents, and other data collected. About 30 percent of the companies that classify their stores based on risk data do so based on a combination of actual loss/shrink and LP measures present in a store, such as EAS, CCTV, and so forth.
Nearly one-half (47.1%) of the participants who assign stores to categories based on risk data have classified between 2 percent and 5 percent of their stores in their highest risk category. About one-quarter (23.6%) of these respondents have classified between 7 percent and 10 percent of their stores in their highest risk category. About 30 percent of these executives report 15 percent or more of their stores have been classified in their highest risk category.
About 30 percent of the participants who assign stores to categories based on risk data report that the average loss/shrink rate (as a percentage of sales) for stores in their highest risk category is between 1.0 percent and 2.1 percent. Almost one-quarter (23.5%) of these respondents indicate that the average loss-shrink rate for stores in their highest risk category is 3.0 percent. Nearly one-half (47.1%) of the participants who assign stores to categories based on risk data indicate that they conduct risk assessments and reclassify stores once a year.
The Loss Prevention Research Council (LPRC) is working to generate more risk and vulnerability rating data and process ideas. Please let us know your willingness to participate in understanding even better ways to measure and predict store and department loss and crime levels.
Read the full column, “Training Makes a Difference,” to learn more about a new online LP/AP problem-solving course through the University of Florida. The original column was published in 2017; this excerpt was updated August 22, 2018.