Between pandemic fallout, rapid adoption of all things omnichannel, and changes in shopper (and shoplifter) behaviors—well, the word “adapt” hardly covers what loss prevention professionals have done these past few years. All facets of conventional loss prevention strategy have had to be reimagined—and now video is poised to take center stage.
To discuss the future of video in loss prevention, Sensormatic Solutions recently hosted a webinar featuring some giants of the loss prevention community:
- Paul Zyskowski, director of engineering at Intel Corporation
- Alex Payne, director of loss prevention at Ingles Markets
- Robert Brown, solutions engineer at Axis Communications
- Brian Field, senior director, the Retail Consulting Practice at Sensormatic Solutions
Let’s take a look at some top takeaways for retailers from the discussion.
The Shift Is On: From Apprehension to Deterrence
A successful loss prevention strategy must focus its efforts on wherever will yield the most results. That used to be apprehension. Now, for a variety of reasons—including changes to laws in many states allowing shoplifters to steal more for lesser charges—deterrence has become the more effective way to mitigate theft.
“It used to be that you go in to catch as many shoplifters as you can,” Payne said. After the dynamics changed these past years, as a loss prevention professional, “you had to do that 180-degree shift. What’s important now is: How do we protect the brand? How do we protect ourselves and our associates? And our customers?”
This shift in posture from apprehension to deterrence for loss prevention puts a spotlight on video as a means to gather actionable data and insights.
“The trend is going to be much more about how to use data more effectively to deter theft at a store,” said Field. “That’s where the struggle is going to be—and where the solutions come into play.”
New Retail Challenge, New Video Purpose
The recent escalation of violence in retail stores through smash-and-grabs, flash mob robberies, and the like has forced yet another challenge under loss prevention—and video’s—purview: safety.
“We’re trying to leverage video surveillance to be more proactive so that a smash-and-grab is not something that you’re always responding to, but rather that you’re able to hopefully predict and prevent from happening in the first place,” said Brown.
Strategic surveillance is key here. The need for visibility hasn’t changed, but where you need that visibility likely has, because omnichannel has changed where transactions are happening and where people are gathering.
Implementing visible video surveillance in the parking lot, for example, provides a visual deterrent for theft or violence before even entering the store. In addition to fixed cameras, body cameras present a real tool for de-escalation in a hostile engagement with an angry customer, Brown noted. “If they see somebody wearing a camera that’s got audio as well, they may rethink how they’re going to express their anger.”
On the Intelligent Edge: Machine Learning and Video
The strides machine learning has made in evaluating video and pictures makes for impressive value on the loss prevention front.
“When you start looking at machine learning, that’s a subset of AI and it’s all about finding patterns in data. And a lot of loss prevention is about understanding patterns,” explained Zyskowski. “Somebody standing in front of an item for a certain length of time versus reaching in faster, or perhaps on Wednesdays, when it’s raining, you have a certain number of items that tend to walk out the door—it’s all these different types of things. Machine learning is about: How do I take and understand the patterns in the data?”
Additionally, Brown noted, utilizing video technology with machine learning gives you a real time benefit because the processing of the visual data occurs on the camera.
“You’re not just looking at pixels changing, which can generate all sorts of false positives,” said Brown. “Instead of drowning in false positives—you’re actually getting a lot closer to real actionable information.”
By utilizing a server in tandem, Brown added, you can share the burden of analysis. “You’re sending metadata or information about the scene to let that more robust solution do a lot more because it doesn’t have to also parse through the video.”
Build Video Solution Buy-in Through a Holistic Approach
When retailers seek to adopt a video solution, they naturally have to overcome budget constraints and margin concerns, and ultimately convince their CFO it’s worth the investment. So, how can you build buy-in? Panelists strongly emphasized using a holistic approach. Video solutions may be typecast as a loss prevention solution, but the truth is the benefits span far beyond shrink mitigation to inventory management, staffing, marketing, merchandising, and more.
“Video leads you to understanding behaviors,” said Field. “And the great thing about that is that now you’ve got behaviors that aren’t just captured on video, but you can turn those into numbers. And those numbers can inform whatever model you need it to inform.”
Brown used the example of four people walking into a store and one receipt to show for it—is that a good conversion rate? Or not? “Being able to differentiate between a family that is shopping together and four people shopping individually is hard to do when all you have is the receipt.”
But when you have video, it allows you to tie data points together—are the shoppers together, where did they go in the store, did they stop at an end cap—and now you have valuable, actionable insights that useful across the enterprise.
Beware the Problem-free Video Solution
The shifting retail landscape coupled with the sophisticated capabilities of retail tech can make a cutting-edge video solution very enticing. It’s new, shiny, and exciting—it’s a must have, right?
Not so fast, the panelists said.
“What we encourage our clients to do is to look for outcome-driven solutions,” Field said. “They have to figure out what they are trying to accomplish. It’s not just simply ‘let’s have some cool tech and throw it out there.'”
Brown similarly cautioned against tech for tech’s sake. “If the reason you want deep learning is because it’s deep learning . . . you haven’t actually identified your problem, and the solution may not be the right fit. It really comes down to working with your loss prevention people, your key stakeholders, and other departments that may have interest in that to say: What challenges are we facing, and can we apply technology to solve those problems?”
For more insight and advice on utilizing video solutions for loss prevention, watch the full webinar. Or, if you’re ready to talk solutions with Sensormatic, reach out today.