Evan Schuman's StorefrontBacktalk

For the purposes of this article, let’s focus on variances in POS transactions and see how we can find theft with an integrated system. To find the theft we would:

• Review the POS data.
• Identify data points that don’t match expectations.
• Identify sales transactions related to that data point.
• Watch the video of these transactions.

As an example, let’s say you operate a QSR restaurant that has an average check of $9.00. You know it is highly unlikely that anyone in the store spends less than $2.00 on any particular visit. We have identified a data point that does not meet expectations (orders under $2.00). You know one of the ways employees may steal from you is under-ringing (ringing up less than the value of the order and pocketing the difference). So you use your video surveillance system to identify all of the transactions for the last week that are under $2.00. One cashier stands out; Billy has nearly four times the number of transactions than the rest of the crew. What next?

This step is where the integration between the POS and the video surveillance becomes critical. The report provides you with a link to the video of that transaction (which has an overlay of the POS transaction). We look at Billy’s less-than-$2.00 transactions, and we immediately notice something. The first order he rang into the POS was for a medium fountain drink ($1.50), but we can clearly see him serving the customer a burger, fries and a shake. When he’s done with the order, we see him take a penny out of the penny slot and move it to the quarter slot in the drawer (to keep track of his take).

Billy is stealing from you. You start getting upset. Then you pull a report in the system to see all of Billy’s transactions for the entire week and realize he has stolen more than $400 through various approaches. You are shocked. You never would have believed it. Billy is a hard worker who is very friendly and always on time. You never would have suspected to even watch the video of Billy because he did such a good job. Even if you had watched a CCTV video of this transaction, you would not have had any reason to believe there was fraud.

Remember, to get the money out of the register without getting caught employees have to game the system in some way. Using these tools, you are identifying situations where there may be foul play. And the integration with the POS is often quite simple. The video surveillance companies typically trick the POS into believing it is a receipt printer and simply “scrape” the transaction data that is fed to the printer and store it in a database to be used later.

That example was an easy one; you came up with the rule yourself (under $2 order). But what if you are not sure what the rule should be? The answer is to look for data that doesn’t match up. Review the number of transactions that had a discount. Does any employee or shift have significantly more discounts than the rest? How about the number of deletes/voids? Does one or more person have more than the others? In technical terms, you are using a crude form of standard deviation to identify transactions that may warrant a closer look. The more advanced tools in this space will actually do the mathematical work for you and identify any transactions that are suspect.

You may be saying “I don’t want my managers in the office looking at video and doing some college math, I want them on the floor.” Would you believe I am aware of at least two companies that will even run the reports, watch the video and E-mail you when they find someone stealing (thereby completely outsourcing your Loss Prevention)? And they provide all of this analysis for a reasonable monthly fee.

Do you really want to drive theft out of your store? Then have the corporate office maintain a rules database in which suspicious transactions are flagged and share all of those rules with every franchisee. If Joe in Mississippi discovers his employees are stealing by not fully activating gift cards, all of the other franchisees automatically get a rule in their system to look for such transactions. (This approach requires that all stores use the same system.)

In my opinion, for stores that are not currently using this technology, it is the single best program that you could implement to improve unit economics, bar none. The results I have seen from this type of system are staggering. I have seen locations realize an increase in sales of more than 20 percent as a result of implementing this type of system. It is one method by which leveraging IT to integrate two data sets (POS and video) can provide enormous value to the business.

What do you think? Love it or hate it, I’d love to gain some additional perspectives. Leave a comment, or E-mail me at Todd.Michaud@FranchiseIT.org.