How to get your business to agree to your security spend? One security man has found a way – have other departments pay for it. He told the Retail Fraud on the Road conference how he’s done it.
Ian Paton, the security solutions manager at building trade product retailer Travis Perkins (TP), wrote in the July issue of Professional Security magazine how TP has gone over to Internet Protocol (IP) to monitor its alarms at its own monitoring centre. He told the event in Leicester that he found it surprisingly easy to justify the spend. He made the point that there have not been many advances in security products; there are still cameras, passive infra-red detectors (PIRs) and gates and so on. TP wanted something different. “We decided about two years ago w would have a vision, moving forward to 2015; and part of that vision was that our only intruder alarm detection was to be by camera.” The reason: using PIRs gives you an alarm, but you cannot see what is happening, such as someone climbing or cutting a fence. To be exact, TP wanted a hybrid: cameras, fences, and PIRs. They found this, through video analytics. “We re fortunate we have our own monitoring centre; we have 1850 branches in the group, and we have about 450 still to move over to a monitoring centre.” The analytics he described as a software programme that takes a picture from the camera and analyses the picture and tells you what is in it. “There’s a big difference between video motion and analytics,” he added, describing video motion as giving an alarm if anything went across and interfered with the picture. “Video analytics makes us smarter. You can tell whether it is a vehicle, human or, very important to TP, a fox or dog; because TP sites outdoors can have an amazing number of activations.” The retailer has had sites, he went on, that could have upwards of 1000 activations, ‘and you think of the man hours that took people to have a look’. That number of false activations at a site went down with analytics to eight or nine – still false alarms, but good by comparison. The alarm directs the operator to where they have to look. You can set the analytics in a car park for example to alarm if someone is inside the car park, and not if a flag is flying or a car is passing on the road outside. The retailer still has physical fences, but with the analytics can set ‘virtual fence lines’ and can set an alarm if a person walks through a gap by walking from one direction (whether, into, or out of, the site). To sum up, with analytics you don’t miss things; and you save on time and money. “I think analytics are the best thing since sliced bread,” Ian said.
Ian passed at this point to Glen Higson, MD of the company Business Insight 3, who admitted that security is still a ‘grudge purchase’, that a company has to do, but wishes they would not – unless you can show return on investment. If IT infrastructure is already in place, what else can we do with it? He offered some ideas; searching for transactions without a customer; an unattended customer; managing f queues; counting people; and staffing levels in a store; of use to merchandising, besides loss prevention, that is. Ian Paton returned to the platform to talk about developing customer intelligence. He recalled how he came into retail 12 years ago and was at Focus DIY. He told a story of a discussion in a store near Christmas; as you walked into the store, there was a pile of drills. Ian warned that if those products were left there, they would ‘fly out the store’; that is, be stolen. The attitude of the ops director however was that Ian was ‘typical security; you stop my sales; I don’t care how many I get stolen’. The ops man’s point; if the desirable product was on open sale, some would be stolen, true; but he would sell many more. In other words, security and loss prevention (LP) have long been seen as people who inhibit sales. Sales and marketing come up with ideas; and security are the ‘bogeyman’, the – to use a Scottish term, as Ian Paton is a Scot – ‘auld enemy’. But through analytics, the LP and sales sides of a retailer are united. Ian gave the example of ‘front end protection’, at checkouts. Especially first thing in the morning and last thing at night, they can be risky places from a LP point of view, as a store can have few customers, or staff. The idea was to get more than one use from every camera and piece of equipment. At point of sale (PoS), you could tie the PoS data from the checkout into the analytics. Every time the till opened, proper procedure says that two people should be there; never only one person opening the till. The analytics software can look if there is only one person present at a transaction, and then send an email to the fraud team. Or, video can be tied into people counting, or sales conversion. Ian Paton said: “I don’t see why in the very near future you count the people in and you could work out the average time people spend in a store; and why shouldn’t we alert the manager on his PC, ‘there are 50 people in store and in another minute’s time you will need another person on a till’.” Ian Paton, then, has looked at this security and LP tool to support marketing and sales.
In what was something of a theme of the day, Ian Paton has also looked at analytics with a view ‘to take guards out of the business’. He said: “They are a very expensive commodity.” He gave the example of three guards at a distribution centre, who might be talking to each other. Analytics can take some duties away from the guards, by doing patrols and let people in and out of a site, perhaps with automatic number plate recognition. So you can reduce the number of site guards by automating, and still running 24-7. “It’s an absolute no-brainer. If I do that to all the distribution sites in Northampton alone I can give the business back more than £1m every year.” He recalled that he did a presentation to the distribution centre manager ‘and their first question was, when can we have it’.
Ian has spoken at previous Retail Fraud events in London each spring. A recent development is a trial in one store to find people loitering in an area with a lot of spit packaging found. An alarm can be sent to the manager’s office who can send a member of staff to offer some customer service to the loiterer. It may be that a retailer is losing power tools or flooring tiles, and the thief is leaving behind the packaging. The store can announce over the public address that a member of staff is called to that particular aisle. The dishonest customer will know he is being watched. The genuine customer will know they are going to get attention; this may lead to a sale. In short, analytics can lead to sales. From the trial in-store, which lead to a rise in sales, the retailer found it had been missing sales opportunities.
Glen Higson returned to round off by suggesting a next step might be facial recognition within a crowd so that not only is someone identified when alone, but within a group of 30 or 40 people, of use for marketing and security purposes. He also aired the possibility of ‘heat maps’, where people are standing longer, so that marketing can see where the high value goods ought to be; and analytics for health and safety, to look for slips and falls; or analytics for tracking of customers’ eyes.
