In the April 2009 issue of Retail Security Magazine we featured a book on data mining – not just for loss prevention but for fraud and crime prevention and detection. Now Barry Vincent puts the case for the benefits of enterprise wide data mining in compensating for dwindling loss prevention budget, how it can effectively reduce losses and bring about a significant change to the culture of loss prevention for retailers.
During times of a downturn in the economy ‘acquisitive‘ crimes such as theft and fraud increase – this is the conclusion according to research carried out by the Home Office during the recession of the early 1990s.
As loss prevention and security professionals we may feel that this is a glimpse of the blindingly obvious, but without entering into a treatise about what makes people commit crime it is fairly safe to assert that in the current economic circumstances some people who suddenly find that their jobs are insecure, their income is potentially reduced or under threat and consequently their normal lifestyle is unsustainable, are more likely to succumb to temptation and exploit opportunities to steal or commit fraud. A further reason for this tendency is that we have emerged from a prolonged period of economic growth fuelling a consumer- driven society where our expectations are higher as we have been used to a regular income, easily attainable credit and increasingly more desirable things to spend our money on. Harold MacMillan’s famous remark that ‘we have never had it so good’ was probably more appropriate to describe the last 15 years than it was at the end of the 1950s!
Reduced resources
It is also an unfortunate reality that despite what the Home Office research tells us about the prospect of an increased threats and crime in the current climate, retail companies will invariably look to reduce costs to accommodate declining sales revenue and falling profits. Unfortunately loss prevention and security often becomes a ‘soft’ target when it comes to making such economies, and as a consequence LP departments find themselves having to restructure or ‘downsize’ to accommodate the impact of a reduced budget at a time when they probably need more, not fewer, resources to tackle the increasing demands of an increase in workload. The inevitable outcome is that loss prevention teams will be required to ‘sweat’ their assets and work harder to achieve their objectives.
Loss prevention will not be alone in feeling the impact of budget cuts at this time. The implications of retailers reducing budgets in other areas of the business are that along with staffing cuts comes the probability that general compliance with operational policies will deteriorate, and people will be working under greater pressure resulting in potential increases in procedural errors and an overall increase in losses due to shrinkage.
These somewhat depressing scenarios points towards an inevitable downward spiral of increasing losses to the business that may only be corrected when the economy turns the corner and budgets can be restored. But does it have to be like this? Does the outcome of a company’s reaction to the economic downturn inevitably have to lead to increased losses, or is there an opportunity to reduce losses through finding ways to enable staff to work smarter rather than harder?
‘Intelligence-led’ approach
In a previous life in the police we also found ourselves confronted with a similar dilemma to that faced by loss prevention personnel in business, that of trying to cope with increasing demands from rising crime with static or reduced resources. In order to improve our ability to address this trend we employed an ‘intelligence-led’ approach. In simple terms this meant collecting data and analysing when, where and how crimes were occurring which then became the basis of an integrated crime information and management system to help better deploy our resources /officers. In the world of retail a similar approach has been championed over the past few years by the ECR Shrinkage Group who developed the method of using the collation and analysis of accurate and relevant data to identify ‘hot’ or high risk products, locations and personnel, and by so doing companies could more effectively focus their resources and activity to reduce shrinkage.
The ECR Group’s research identified four specific areas where shrinkage occurs, supplier fraud; process failure; external customer theft and internal staff theft. Their research identified that theft and fraud make up about 75 per cent of total shrinkage, whilst failure in processes eg compliance errors or failure to correctly manage the handling and sale of products, made up the rest. The ECR group recommended that mapping and measuring risks and identifying root causes of loss in these areas, with a systematic approach to gathering data as part of a process was essential for companies to be truly effective in reducing losses. The success of this approach depends on a number of factors, including a company-wide commitment to loss prevention, using the right products etc, but equally critically to deploy proper measurement of their losses and collect credible and accurate data.
Point of sale monitoring
The ECR Road Map approach to shrinkage is now widely accepted, and retailers have adopted its recommendations and introduced a range of initiatives to reduce shrinkage. Amongst these include greater collaboration with suppliers to improve product security; significant investment in security hardware, including product protection measures in stores; better store inventory processes and the adoption of computer based technology particularly for Point of Sale (PoS) monitoring. Using analysis of POS data has helped to detect both operational errors e.g. pricing and keying mistakes, as well as potential dishonesty by staff during till transactions.
There is no doubt that for most retailers PoS monitoring has become a key tool in reducing loss by inhibiting staff dishonesty, providing a forensic trail for detecting fraud at the tills and catching dishonest cashiers, and especially shedding light on the practice of ‘sweethearting’ (collusion between staff and customers), the true extent of which is largely uncertain. However, for many retailers PoS monitoring has become synonymous with data mining although its impact is limited to ‘front end’ data and activity, and it does little to address those areas of loss from shrinkage further back along the supply chain where operational processes and people’s activities are not subjected to the same level of scrutiny and where losses may be greater than at point of sale. In general these areas of the business are often difficult for loss prevention teams to penetrate and few meaningful data about loss generating activity is available.
Enterprise-wide solutions
Arguably, it is in these areas of the supply chain (from supplier to the backroom of the store) where the opportunities for staff theft and fraud, or the lack of compliance with operational policies that contribute significantly to shrinkage, are far greater than at point of sale. Supply chain and distribution networks are often seen as so complex and operating with such a high tempo to meet ‘just in time’ requirements that in many cases only lip service is paid to the processes and procedures designed to ensure the security and integrity of the products passing through them to stores. In such an environment efforts to maintain the disciplines of effective security often go unenforced by the operators, so that for example the integrity of security seals on vehicles, and the checking of deliveries from suppliers in distribution centres among other measures go by the board. In these circumstances loss prevention tends to becomes loss reaction, where only the significant incidents are reported and followed up and few data collected that could help to limit losses consistently.
How many companies collate the information that can identify their best or worst performing suppliers, beyond their capability to deliver on time? How many identify those individuals in their own distribution centres who make most or fewest mistakes in goods receiving so that better supervision can be given to some, and better performing staff be matched to receive suppliers who’s deliveries are problematic? In my experience few companies have data mining capability in the key areas of their supply chain especially delivering and receiving and in their wider logistics and warehousing functions.
It is even less likely that companies will have visibility of loss data in another critical area of their business, that of the purchasing activities of their buyers. It is here where inappropriate and unregulated relationships with suppliers can lead to significant and costly fraudulent activity. Generally problems in this are only brought to light by a ‘whistleblower’ rather than through effective day to day management using appropriate monitoring tools. Why do we think that only checkout operators are capable of making errors or committing fraud, and therefore warrant close monitoring when there are just as many, if not more opportunities, for buyers and suppliers to commit fraud with significantly greater sums of money involved?
Such enterprise wide solutions to reducing loss through data mining are not commonly deployed in the UK, and many companies are reluctant to consider introducing them because of the potential disruption to their IT systems or the perceived high costs involved in developing such a solution when the Return on Investment (Roi) is uncertain. However, there are a number of innovative ‘off the shelf’ data mining solutions available that have very minimal impact on existing IT systems, and that at the same time can achieve a swift and significant ROI.
Getting the grease to the squeak
Surely, at a time when loss prevention teams are bracing themselves for a difficult and demanding few years and retailers generally are tightening their belts, an investment in data mining across the breadth of their whole operation rather than just at point of sale offers a real opportunity to make a significant impact on shrinkage. Having a system in place that gathers and analyses accurate information about loss generating activities, incorporating not just exception reporting but also the issuing of alerts, together with guidance on the relevant corrective action is the key to getting the ‘grease to the squeak’, providing managers with the capability for effective intervention to stem losses in a timely and appropriate manner.
Now is the best time for companies to invest in the more extensive deployment of data mining tools that will not only help to them make optimal use of limited, and in some cases, declining loss prevention resources but also to develop a genuine business -wide culture of loss reduction where managers at all levels not only share the accountability for reducing losses but have the necessary tools to reduce theft, fraud or error across the business and significantly impacting the company’s bottom line.
About the wrtier
Barry Vincent is a security consultant and former head of security for a retailer’s distribution network. He is also a non-executive director for Intrepid Security, the UK partner and distributor for Profitect, featured in the April issue of Retail Security Magazine, who offer enterprise data mining software.
