Many Loss Prevention Professionals are Missing the Big Picture – and Significant Sources – of “Total Retail Loss” Because They’re Solely Focused on the Point of Sale
If you’re in loss prevention, you’re no stranger to working under pressure. Specifically, I mean the pressure of doing more with less, including times when corporate increases its expectations of your team with no corresponding increase in budget or other resources.
When those mandates come down from above, where do you turn first? Probably to the point of sale (POS), which is smart. As the point at which money changes hands, the POS is a rich source of evidence of cash loss. Your second problem area might be the returns desk, right? That makes sense given this is a common target for organized retail crime.
But it’s not the only source you should be focusing on, by any means. What about your store employees’ inventory practices? We all know plenty of loss happens via seemingly harmless inventory issues, but it’s hard to justify allocating resources to them when the POS is often seen as a priority.
So, how can you monitor these subtle, yet costly inventory issues without losing focus on the spot where cash is most vulnerable?
With technology, that’s how!
I always recommend loss prevention teams invest in artificial intelligence (AI)-powered analytics solutions, like prescriptive analytics. Prescriptive analytics can comb through inventory data and both identify any profit leaks and send course-corrective actions to the right person for prompt resolution.
Here are several mundane inventory practices that AI-powered analytics can monitor for loss while you focus on the front-end theft, fraud, and error-caused losses at the POS or returns desk.
Adjustments
Adjustments are a perfect example of a routine inventory practice that can easily contain evidence of fraud. Scores of adjustments happen at the average retailer every day, many perfectly normal. Data-entry errors, under- or over-deliveries and damaged product all require an adjustment to keep store records as accurate as possible.
But you don’t need reminding that adjustments can easily conceal losses. Perhaps a careless employee made a fat-fingering error on an adjustment, accidentally inflating stock by thousands of units and making out-of-stocks inevitable. Worse, perhaps another employee flat-out stole product from the back room and adjusted stock levels to cover up the theft (i.e., a manager stole two e-cigarettes off the shelf and negatively adjusted the store’s stock level by two so no one would miss them during cycle counts). Trying to find inappropriate adjustments in a sea of countless legitimate ones is a time-consuming process – time that any executive could argue would be better-spent focusing on the POS.
The right AI-powered analytics solution can help. Numerous large retailers have their solutions configured to monitor inventory adjustments for suspicious activities. Examples of these include:
- Adjustments on known theft targets immediately after cycle counting (the perpetrator may have waited to steal until immediately after the count, hoping the loss would be seen as a counting error)
- Adjustments that are above average in monetary value, frequency or quantity
AI-powered analytics solutions immediately alert the relevant person to such findings and send them course-corrective actions advising what to do next.
In one notable case, the Zebra Prescriptive Analytics Inventory module advised a specialty retailer’s asset protection (AP) team that a warehouse had made a negative adjustment on a brand of diapers amounting to nearly 100,000 units (almost three million diapers). The warehouse manager told the responding AP investigator that the adjustment was perfectly normal, as the ongoing pandemic had created supply-chain disruptions that made huge inventory adjustments routine. However, ZPA identified that this particular adjustment was exponentially above average for the warehouse and the product line – evidence the investigator leveraged to insist the warehouse manager check stock levels anyway.
It turned out the adjustment should have been for 10,000 units of diapers that the manufacturer never delivered, but a fat-fingering error had accidentally added an extra zero to the adjustment. Thanks to the prompt alert of the analytics solution, the retailer avoided any wasted replenishment spend this error would have caused, had it gone undetected.
Store transfers
Traditionally, many retailers executed store transfers only in cases of extreme need – until, that is, COVID-19 supply disruptions made them commonplace. While transfers were most helpful in balancing scarce product across locations, it also exposed retailers to numerous opportunities for loss in the process.
There are many ways store transfers can go horribly wrong, including:
- Transferred products get lost in transit or delivered to the wrong store
- Store A employee scans the right quantity of items to transfer, but doesn’t pack all of them
- Store A employee forgets to adjust their store’s inventory after executing transfer
- Transfer system may put the transferred items directly into store B’s inventory before they’ve even left store A, making store B file the discrepancy as shrink
- General employee makes mistakes due to unfamiliarity with the transfer process
- Store B employee is unaware of the transfer, assumes it’s a mis-shipment and puts the whole quantity on markdown, or sends it back
All of the above hiccups are very subtle and time-consuming to identify without AI-powered analytics – because, as you may have noticed, all of them can easily be false alarms. With AI-powered analytics, you can configure a solution to monitor all data transferred between stores. The right solution can easily identify gaps in transfer processes that cause total retail loss and send course-corrective actions to the person best-suited to intervene.
Miscategorized damages
Damages are inevitable in retail – sometimes avoidable, sometimes not. Most, if not all, store employees have full authority to label products as damaged within their stores’ inventory management systems. This may make perfect sense on paper, but it also exposes you to significant risk.
Zebra customers have encountered numerous cases of employees using the “damaged” category like a “miscellaneous” folder on their computer or a junk drawer in their house. When rushed or confused, new or undertrained associates may scan products as damaged simply because they don’t know what else to do with them. Needless to say, damaged products are unsellable and must be destroyed, making total retail loss inevitable.
Does sifting through thousands and thousands of daily damages to find the improper ones sound impossible? That’s because it pretty much is, for the typical AP investigator. But AI-powered analytics, like prescriptive analytics, can be configured to monitor damage scans for inconsistencies and point you directly to any problems.
The methodology is quite simple: the solution analyzes damages at a variety of levels (i.e., by store, district, product, etc.) and determines benchmark averages. Once it has those benchmarks, it can easily identify above-average damage activity, whether in terms of monetary value, frequency, or something else.
In one example, a fashion retailer’s prescriptive analytics solution alerted an AP manager to a store that had scanned in $11,000 worth of damages in a single day and directed him to do a compliance check. It turned out the “damages” were not damaged at all – they were online purchases that had been returned in store. The two associates on duty had never encountered this before and, thus, scanned them all in as damages. Thanks to the fast prescriptive alert and course-corrective actions, the AP investigator was able to intercept the perfectly saleable products (en route to a destruction facility!) and return them to the shelves.
There’s Just One Caveat
It’s important to remember that this prescriptive approach is only as good as your inventory data. If you’re missing information or the data in your inventory system is outdated, your loss investigations are going to be skewed and prescriptive actions could be misguided.
This fact alone makes self-directed inventory count solutions another worthwhile investment. You can now download software onto associates’ mobile computers so they can conduct all inventory counts in house, on your terms and schedule. Technically, cycle counts could occur every day if you have a team member or two who can focus on them. You don’t have to rely on annual wall-to-wall physical count data anymore – or hope that a third-party service provider will show up on time to conduct the count.
No matter what you do, though, just be sure you don’t get so caught up on eliminating the most obvious sources of loss that you overlook the inventory issues that could lead to equally significant losses long term.
Let’s chat if you think it’s time to look at – and reduce – total retail loss in a more prescriptive way in your stores.
###
Related Reads:
Suresh Menon
Suresh Menon is Senior Vice President and General Manager for Software Solutions at Zebra Technologies. He is responsible for the strategy, development, and management of Zebra’s Software Solutions product line, which features an industry leading portfolio of AI-powered workforce management, task execution and communication solutions, prescriptive analytics and self-directed physical inventory management solutions.
He has held a range of software leadership roles, most recently serving as the Senior Vice President and General Manager of Informatica’s Master Data Management and 360 Solutions business unit where he led the transformation of an on-premise, perpetual licensed business to a high growth, cloud native and subscription business. Under his management, Informatica’s Master Data Management and 360 Solutions were recognized by Gartner in 2020 in the Leaders section of their Magic Quadrant as the leading solutions in this category.
Earlier in his career, Suresh served in software product management leadership positions at Identity Systems, a division of Nokia Enterprise Solutions and at Search Software America.
