Skip to main content

Focal

From Weeks of Footage to One Hour: How Shelf Data and CCTV Teams Work a Theft Case Together

A guy walks into a grocery store, heads to the candy aisle, fills a bag and leaves. Nobody notices. The product never touches the register, and a couple of days later someone on the store team adjusts it out of inventory. In most stores that would be the end of it. The loss gets rolled into the shrink figure at the end of the period and nobody ever finds out who took it or when.

This time it went another way. Shelf data from Focal pointed the loss prevention team to the aisle and the hour the candy disappeared. They pulled up their security footage, found him, and shared his description with the store team. He came back the next evening, did the same thing, and the store was waiting for him at the door.

Focal didn’t catch him. The store’s security team did. What we gave them was a place to start looking, and that turned out to matter more than you’d think.

How a theft alert gets triggered

Focal cameras take pictures of the shelves throughout the day. Theft Spotter lines those pictures up against the store’s sales and inventory data, which comes in every hour.

If a product is on the shelf in one image and gone in the next, and nothing sold in between, that’s a flag. If the store then adjusts those units out of inventory, usually within a few days, the flag becomes an alert.

The alert goes to the loss prevention team with the item, where it sits in the store, and the hour it went missing. They can also see the shelf images from before and after, plus the sales history for that item.

What Focal doesn’t do

People mix this up a lot, especially when facial recognition comes up in the same conversation.

Focal doesn’t stream video. We work from still images of the shelf, and our alerts show up after a loss is confirmed, not while it’s happening. Our cameras are aimed at products, not people. In fact, we drop the pixels completely so no person ends up in an image at all. We have no way of knowing who took anything.

We also don’t send images to facial recognition tools or watch lists. If a store uses those, that’s their system, run by their own security staff under their own policies. Our part ends at telling them which shelf and which hour.

Handing it over to the security office

Loss prevention runs separately from store operations. Their offices are usually off limits to most store staff, and a big part of the job is sitting in front of monitors scrubbing through video.

Without a lead, looking into one missing product can mean going through days or weeks of footage across several cameras. Most of the time nobody has that kind of time, so the loss never gets looked at.

An alert changes that. The team knows something like: this item left this shelf on Thursday between 10:55 and 11:55. They check the shelf images and sales to make sure it wasn’t sold, open their own video system, pull up the cameras on that aisle for that hour, and watch until the product leaves the shelf. Then they follow whoever took it through the store. If it’s theft, they handle it the way they normally would, whether that means building a case, alerting store staff, or passing details to other stores.

Everything to do with identifying people stays with them. We just keep them from searching blind.

Why an after-the-fact alert still pays off

By the time an alert comes through, the product is gone, and that particular loss usually isn’t coming back. But people who steal from a store tend to come back too, often to the same aisle.

It gets more useful across a group of stores. When the same items keep disappearing from the same aisles at several locations, the pattern is hard to miss. For organized retail crime, where the same crews hit store after store, that gives loss prevention teams something concrete to compare and act on.

It also gives loss prevention something they’ve often been missing: steady, specific leads. Store staff are usually told not to confront shoplifters, so without those leads LP has little to go on beyond catching someone in the act.

Sometimes it isn’t theft at all

Some alerts turn out to be process problems once the team looks at the footage, and those are worth finding too.

A product won’t scan, so the cashier rings it up under another item. The shelf empties, the correct item shows zero sales, and it looks like it walked out the door. We’ve seen a fresh item that was fully stocked every morning show no sales at all, because cashiers were keying it in as something else. And when the deli or kitchen pulls product off the shelf for its own use without doing the transfer, the system just counts it as missing.

None of that is theft, but all of it ends up in the shrink number. Sorting it out tells store leaders which losses need security and which need some training.

Shelf data and security footage work better together

The shelf can tell you what’s missing and when it went. The camera in the ceiling can tell you who took it. Neither gets you very far alone, but put them together and a vague shrink figure turns into cases a team can actually work.

We keep our part of this narrow on purpose. Focal watches the shelves, keeps shoppers out of the picture, and gives loss prevention a starting point. They take it from there.

Want to see how Theft Spotter fits with your loss prevention team? Let’s talk. 

Frequently asked questions

Does Focal use facial recognition?

No. Our cameras capture shelves, and we blur out anyone who shows up in an image. We don’t identify shoppers or send images to any identification system.

Are Theft Spotter alerts real-time?

No. They’re built from shelf images, hourly sales and inventory data, and the store’s own inventory adjustment, so they come in once a loss is confirmed. The upside is that security gets an exact hour to review.

How do loss prevention teams use the alerts?

They take the item, shelf location and time window and go straight to that footage in their own video system, instead of searching through days of video.

Can Theft Spotter tell theft apart from operational errors?

The alert flags a loss with no matching sale. Once the team reviews the footage, plenty of cases turn out to be scanning or process mistakes, which helps stores decide where training is needed.

Written by:

Let's talk!

We can custom-build a solution just for you.

"*" indicates required fields

You Might Also Like