Applications
Boots remedies till fraud with data mining
Real-time alerts warn store managers
By Leo King | Published 13:39, 06 February 08
Boots has saved “millions of pounds” since hooking up a point-of-sale data mining system in 2006, it said this week. The system delivers real-time alerts to store managers so they can act quickly to stop suspicious transactions.
The high street pharmacy chain, which has 2,300 stores and employs 60,000 people, told delegates at the Retail Business Show in London that it had decided to implement the system after discovering numerous attempts by a minority of staff to undercharge for items or to refund excessive amounts of cash for goods they had bought more cheaply. Boots processes 1.5 million transactions each day, and with sixteen different ways to pay, including loyalty cards and vouchers, it had become difficult to track the problem.
Also in this channel
Till fraud is common in high street retail, and other stores such as Monsoon have spoken out about the need to identify the problems as they occur using sophisticated electronic point of sale systems feeding into business intelligence software.
Robert Jennings, head of loss prevention at Boots, said the chain had installed PCMS Vision Smart Store two years ago, linking into a March Networks data mining system, in a “radical” attempt to eradicate the problem.
“We can now deliver simple, real time alerts to store managers that they can follow, rather than the usual over-complex business intelligence reports. This means they can intervene early,” said Jennings. Currently 5,500 alerts are delivered every day.
In one instance of fraud, Jennings said, two consecutive refunds of £250 were made on a till, linked to the same loyalty card but without any purchase having been made. The employee was later dismissed and arrested by the police. In another case, an employee discounted the price of a baby monitor from £140 to £10, bought it and refunded it at full price in another store. The employee was found to have three loyalty cards linked to other “suspicious” sales at Boots.
“We also generate daily reports that are measured against 150 markers of suspicious transactions, and we can update those markers when we want,” he added. “We have saved millions.”
Jenning said it was vital for businesses using till data mining to engage all levels of staff, and for a reporting culture to be encouraged.
Boots now plans to integrate suspicious behaviour alerts from closed circuit television cameras with the point of sale alerts system, to further improve the quality of crime warnings.
