How Spar tackled loss prevention
Loss prevention is a significant problem for retailers and statistics show it’s on the rise. For Ravi, the owner of this Spar outlet, stock loss represented over $6,000 per month.
Working with Verifyfaces, Ravi installed a facial matching security system and has seen theft reduce by 40% to date.
With the cost of goods increasing and margins shrinking, stock loss from theft can be devastating to the bottom line.
This was the case for Ravi, who operates a busy Spar retail store in Parramatta. He was seeing losses of at least $1,500 each week. He contacted Verifyfaces and agreed to host a pilot of their facial matching system so he could identify and monitor those most likely to steal from the premises. Shoplifters target busy stores, attacking regularly and often working in groups.
This unique system allows Ravi to identify and tag suspect individuals from video footage. Cameras are carefully placed to capture anyone entering the premises. The facial matching system searches the image database and confirms a match. An alert is sent to the manager who approaches the individual and ‘offers assistance’, subtly letting him know he is being monitored. The Verifyfaces system does not store any data on the general public, it only records data on people that are known offenders.
The system uses the same face matching algorithm that is trusted by border force authorities and police organisation around the world. It’s been used and refined over 40 years in the field.
The system has been in place for 18 months and losses have decreased by 40% with further improvements expected.
Initially the Verifyfaces system was being operated using onsite servers, however the cost of this hardware was prohibitive for most retailers. In addition, latency issues also affected system performance at times. So, Verifyfaces partnered with Eagle Eye Networks to developed a cloud based solution that overcomes the performance issues and greatly reduces the overall cost of the system.
Searching surveillance footage is just like searching the web. Users don’t have to manually review footage, they simply enter instructions, for example, “search for male with a red shirt, or woman with a backpack”. These analytics allow the operator to locate and track persons of interest across all cameras and instantly find what they are looking for.
A Reseller will be in touch with you within 48 hours.