For generations, the retail risk mitigation solution portfolio has remained virtually dormant. The same methodologies of the 1970s are still painfully evident today—perhaps in a different form factor, but with the same precise intent: limiting exposure to malicious intent losses, such as shoplifting.
Locking showcases, electronic article surveillance systems (EAS), alarms, acrylic keeper cases, unmonitored CCTV systems, prosecutorial warning signage, mirrors, and numerous restrictive fixtures and gadgets on at-risk merchandise are all manual, disconnected, unintelligent, and singularly focused on external theft.
None of these addresses the other, and arguably larger, percentage of total loss happening upstream—administrative and process losses. These include those incurred at any point from the retailer generating and transmitting a purchase order, order fulfillment, supply chain and warehousing, and ultimately delivery to the retailer’s receiving docks.
Risk begins the instant a purchase order is transmitted—but almost exclusively, the focus on mitigating loss remains within the four walls of retail stores, and those losses are only revealed after conducting an inventory. This gross mischaracterization—that “it’s all theft”—has dominated the industry for decades.
The surge in retail losses during and immediately after the pandemic—with countless social media clips of smash-and
-grabs, car chases, and physical altercations—led many to conclude that theft was the sole or primary source of retail shrink. That’s a dated assumption not supported by reality.
The Silos of Retail: Not All Retail Channels Are Equal
As store operations and all other facets of retail operations have evolved both procedurally and technologically, they’ve left behind the attribute of financial risk and losses.
Just in the last ten years, retail store formats have changed immeasurably thanks to holograms for merchandising, computer vision, robotics, radio frequency identification (RFID) for inventory management, numerous options for the customer to shop and pay, and a multitude of delivery options (in-store, BOPIS, e-commerce, home delivery, and drones).
Stagnant until the 1990s, the toolkit of retail loss prevention professionals was limited. If any improvements were evident, they were incremental in value.
Data analytics with some point of sale (POS) integration strategies, albeit rudimentary by today’s comparison, began to emerge, but logically, those resources were completely dependent on the source and accuracy of the data.
Not unlike any other profession, one key and undeniable factor influencing the retail LP industry is the talent and leadership over the years, more specifically, the quality, experience, business acumen, and focus.
Understandably, in decades past, the heaviest density of talent in the retail LP career field was former law enforcement and those with a keen sense of nefarious intent, suspicious behaviors, and a real passion for the mission of deterring criminal intent in their respective stores.
As retail operations have become more complex and sophisticated, so too have the demands of the LP industry. A shift became evident around the turn of the century with more operationally minded leaders taking over their company’s loss prevention teams, ultimately leading to a shift in focus from almost exclusively tackling shoplifters to include the more encompassing total store operational and process efficiency gaps, now commonly referred to as “total retail loss.”
The early 2000s ushered in the age of digital options advancing the earlier versions of data analytics, and also saw the emergence of RFID technology, initially almost myopically focused on supply chain and inventory accuracy.
As the maturity of the solution portfolio continued to evolve, the 2010s introduced smart versions of earlier technologies—systems which could provide root cause, suggest improvements in procedure, and, to some degree, predict loss opportunities.
A great example of a customer-facing technology introduced with little to no forethought of a risk mitigation strategy is the headline grabbing self-checkout kiosks. They began to emerge in the late 1980s with varying degrees of adoption, and until about 2014—identifying and limiting the now-proven significant loss opportunities—there were no effective technologies to mitigate. Sure, there were a handful of good old college tries, such as weight scales, signage, public view monitors, and limitations on items, but none that really made a dent.
Decades passed with near-complete exposure to losses from self-checkouts. Otherwise, a handy payroll reduction for the retailer, the exposure to risk was essentially ignored or marginalized by retail leadership until the mid-2010s, with the advent and scalability of computer vision.
All of that said, it has been a slow dance from the 1970s, when no visible changes in merchandise protection were introduced, until the early 2000s. That means over three decades of routine, repeating strategies, all while the other facets of retail continued to expand, explore, test, trial, experiment, and otherwise push the envelope.
The LP industry’s journey from the 1970s where the focus was more physical security and gadgets, to the ’80s where retail began to see hints of technology in EAS and CCTV systems, to the ’90s offering early POS analytics, to the turn of the century with RFID and the 2010s introducing “smart” solutions—has assuredly been a slow one, but improvements nonetheless.
Although the loss prevention industry is often considered the retail technology laggard, it would be inaccurate and unfair to say it is the fault or lack of focus of LP leadership. Historically, it was more about the ability to wrestle the limited and often anecdotal data to prove both the losses and risks and, more importantly, to build a defensible ROI table for meaningful capital investments designed to mitigate these losses—much less to secure research and development capital for next generation solutions.
As the then-US Secretary of Defense, Donald Rumsfeld, famously said in 2004, “As you know, you go to war with the army you have, not the army you might want or wish to have at a later time.” Thus is the daily life of a retail LP executive.
Loss and Risk Data Ambiguity: A Real Challenge
Of all the petabytes of data most retailers have at their fingertips today, the one segment of retail that is short of specificity on loss data is LP.
Of course, retailers are aware of and can quantify instances of shoplifting and the results of internal investigations (presumably, those who were caught), but from there, the tabulation of loss becomes very murky. Most have no idea where in the value chain losses occur, whether at the supplier, during transport, due to administrative errors, their distribution network, or at the store—it’s a guess. Granted, it’s likely a sophisticated guess based on best available data, but still a guess. Possibly, “estimation” is a more suitable term.
This challenge with data ambiguity does not exist elsewhere in a retail organization. Retailers have easy access to data on the number of items ordered, billed, received, sold, and returned. But loss data is the black hole; it is just absent.
Having true financial item-level shrink still eludes most retailers. When the annual inventory is completed and the results provided, the store knows their total store shrink, and possibly down to a department level, but subordinate to department level, there is no identified financial shrink at item-level. In other words, a retailer can’t accurately and definitively report the financial shrink of each item in their inventory in the same way they know sales performance and margins.
Even if item-level financial loss data were available and accurately reported, the “how” remains the elusive element of the question. Where did the merchandise disappear? Was it a keying error somewhere? Was it even shipped from the supplier? Did it arrive at the distribution network? Can they confirm physical receipt or just view the invoice at case/pallet count level (not at the item level)? Did the store even receive the merchandise? Thus is the challenge in accurately quantifying the actual losses.
Tragically, when that inevitable shrink number is reported after the cyclical inventory, the all-too-often reactive assertion is that it happened at the store, and it was likely theft. I heard this countless times during my years in retail. Theft was always the laundry list leader of “how did this happen?” It’s a dated assertion, but one often heard. The fact is, there is no definitive answer.
The Technology Inflection Point: The Change
The risk mitigation solution set readily available to retailers to combat risks both physical and digital really began to change in the 2010s.
A few promising technological advancements stand out above most, which, of course, include the savior (sarcasm intended)—artificial intelligence, more specifically computer vision, and the now more broadly deployed RFID technology in the use case of shrink visibility, the benefit of which is item-level accountability. No more guesswork of whether the retailer received the item, where it is, where it isn’t, and, germane to this conversation, to provide light in the dark room of the sources of shrink.
The combination of these two technologies alone is an almost unmatched force multiplier in the retail risk mitigation space.
Not only can retail now “see” everything, they can also gain intelligence down to the item-level, virtually eliminating the long-standing anchor of ambiguity in sources of loss.
With this data, retail LP leaders can more surgically address loss opportunities, be they upstream or in the store, where the honest customer is not penalized for a supposed or estimated problem.
Technologies exist today where, in real time, suspicious instances can be reported to the store team to intervene, provide customer service, and subliminally deter attempted theft or other malicious activity. No longer are the days of gumshoe tactics the sole option.
Culture Shift: Where Retail Silos Dissolve
The silos of retail have been coming down over the last ten years or so, where technologies initially focused on core retail operations can be easily adapted to include LP use cases. Gone are the days of the isolated, one-trick-pony solutions where the ROI is singularly theft.
This amalgamation matters in innumerable ways, from capital allocation to customer experience, to intelligence collection and sharing, to the black hole of specificity of source(s) of loss.
Now What?
Computer vision and RFID are two force multipliers mentioned earlier, with profound combined value in both the store operations and risk mitigation spaces, end-to-end.
These aren’t emerging, shiny, and therefore cool objects in the retail space. In the case of RFID, it has existed in retail since the late ’90s when the MIT AutoID Center, under the leadership of Dr. Sanjay Sarma, formally introduced the technology to retail almost exclusively around supply chain and inventory management. The value extension to include item-level shrink visibility came years later and, to this day, struggles to find broadly accepted and understood value—which is perplexing, to say the least.
While still evolving, computer vision affords a host of values to retail, from inventory management to sales, to process control, to risk mitigation, and beyond.
As many are aware, AI is ubiquitous, with a perceived unlimited range of applications. Go to any trade show, and if there are one hundred exhibitors, eighty of them will showcase AI, whether real or hype. As with any other promising technology, and in the light of the Gartner Hype Cycle, things will settle down, and the real “so what” of AI will continue to come into focus. Retail will be a major part of the defining applications of AI. In fact, that’s already in motion.
It Is Now a Choice
Around the early 2010s, the available solution set was quite limited. The opportunity to engage or explore with non-traditional solutions was a more intentional and nuanced endeavor requiring a mindset of “What we’re doing today isn’t sustainable. Something has to change.” Retail LP leaders didn’t have many options beyond the legacy kit.
This is not the case today. Leaders in these professions have more choices than ever before, and all with value propositions that encompass various facets of the retail enterprise. No longer are they left with the one-trick ponies of yesteryear. They’re “at the table” in a meaningful and fresh way.
They can now join, actively engage with, and even lead more sophisticated technology conversations where drones, AI, computer vision, robotics, and even holograms are discussed, considered, and possibly pursued. All that is needed is a passion for change and appreciation for what these newer technologies can do in the risk mitigation space—and the list is long.
The silos have faded, and this has opened the doors to more innovative, customer-friendly risk mitigation technologies.
It is now a choice for a retailer to either continue the legacy solution set of EAS systems, merchandise tethers, law enforcement/security presence, customer flow gates, restrictive fixturing, locking showcases, and the like, and expect better results, or to join the existing technology streams of their operations counterparts and put the computer vision and RFID technologies to work in their space too, with a far more profound and specific impact.
It is now a choice how high-risk merchandise is protected and offered for sale.
It is now a choice to remain uncertain of the sources of loss and to estimate what needs protection on the sales floor, negatively impacting the shopping experience of all customers.
Exasperated by reduced store staffing? It is too easy to quickly order online and have items delivered to your home, possibly on the same day, than to wait however long for assistance with locked-up items.
It is now a choice to not use computer vision to provide advanced warning of a potential active shooter in the parking lot, of a sweeping theft or organized retail crime event in progress, of a potential slip/trip fall, an assault or child abduction, an arson event, a skip scan at the self-checkout, or any risk routinely experienced in today’s retail world. These are choices.
Computer vision doesn’t take a break. It doesn’t update its social media account while on the clock. It doesn’t bury its head while roaming the sales floor, walking past customers, and failing to provide any opportunity for customer service. AI models continuously “learn” from images, improving and becoming more and more accurate. Computer vision can “watch” multiple feeds and monitor and alert either in near or real time, likely before the human even knows something happened.
What’s Ahead?
One of the most compelling emerging technologies is the use of holograms. Already present in major airports, assisting travelers with gate directions and flight information, holograms are now appearing in retail, particularly in marketing and advertising applications. Scroll through social media and you’ll likely see video clips of these displays capturing attention and sparking curiosity.
As a result of the pandemic, many retailers significantly reduced labor hours. Post-pandemic, most have maintained those reductions, realizing they could still meet performance targets. The result: a near-anemic customer service presence on sales floors. Today, most shoppers turn to their phones for answers rather than wait for an employee, especially when faced with locked showcases.
Imagine if holograms could provide virtual customer service in high-service areas, powered either by AI or by a remotely located human via live stream. The result: instant engagement, improved service, and likely increased sales.
But there’s another benefit: deterrence. LP professionals know that bad actors prefer privacy. They don’t want to be seen or interrupted. The presence of a hologram—even a nonhuman one—can alter that comfort zone.
Many will recall the classic line from Star Trek: “Beam me up.” At times, the show depicted life-size holograms appearing out of thin air. Now imagine such a technology placed in high-theft areas. The visual impact alone could give would-be thieves pause. Honest customers might be intrigued, delighted, and more likely to engage. But those with ill intent may react quite differently.
They might wonder: Can this hologram record me? Identify my face? Capture my voice? Follow me out? Read my license plate? Even if those capabilities don’t yet exist in that exact format, the perception of being observed may be powerful enough to alter behavior.
The retailer, in this case, avoids loss, protects human staff, and leaves a lasting impression—on both customers and criminals.
This is not science fiction. It’s not fully scalable—yet. But the groundwork is being laid. And its implications for customer experience and risk mitigation are profound.
Got It. Now What?
Innovation is endless; it does not pause so that others can catch up. In many cases, the pace of technological advancement has outstripped retail’s ability to adopt and implement. All that’s required is thoughtful investment and an unrestrained imagination.
The challenge for today’s LP teams is to proactively partner with their operational, technology, and innovation counterparts—to integrate efforts, align objectives, and broaden the scope of what technologies like RFID and computer vision can deliver across the enterprise. LP must articulate the broader value proposition, not just in theft deterrence but in overall risk reduction, operational efficiency, and customer experience.
Today’s LP leaders should demand innovation. They should expect transparency and collaboration from technology providers—or be willing to walk away. The old strategy of slow, incremental improvements is no longer effective in today’s volatile retail environment.
Loss prevention leaders must step outside traditional vendor pools and explore partnerships with non-traditional sources: defense contractors, research institutions, universities, and think tanks. These often-overlooked communities may be developing breakthrough technologies with untapped applications for retail.
Engage them. Share the pain points. Build meaningful, two-way relationships.
The retail asset protection profession has changed dramatically over the course of my thirty years in the business—mostly for the better. We’ve had struggles, setbacks, and successes. But today, we are no longer waiting for innovation to reach us. Innovative solutions exist. They are real. They are viable. They are increasingly affordable.
We have choices. Have we reached our Sputnik moment? Are we there yet?
Brand Elverston joined Walmart in 1995 after serving over eleven years as a Field Artillery Officer in the United States Army. He began his new retail career with the field Asset Protection team but quickly transitioned to the home office. For the last fourteen months of his nearly 22-year career with Walmart, Brand was appointed the Director of the newly created Operational Execution team. Brand retired from Walmart in March of 2017 and is currently the Founder/Principal of Elverston, LLC, is an Executive-in-Residence (EiR) and an advisory board member with the Auburn University RFID Lab, a member of the University of Alabama College of Arts & Sciences Leadership Board, and serves on the Advisory Council with Graham Partners private equity. Brand has served on several other committees and boards including the Retail Industry Leaders Association (RILA) Asset Protection Steering Committee, Chair of the RILA Horizons Technology Committee, and served as Co-Chair of the GS1/EPCglobal EAS/RFID joint requirements group.