Loss Prevention: Managing through the “Perfect Storm” 

Retail loss prevention is potentially facing a “perfect storm.” After years of steady shrink, the industry is expecting to report the biggest growth in shrink losses in more than a decade for 2008. With deteriorating economic conditions, 2009 is expected to be even worse. The increases are coming when, after years of cost cutting, retail LP departments have their lowest headcount in two decades. How can LP executives manage to this impending crisis?

The most proven answer to managing increased risk with fewer resources lies in technology. So, can LP rely on current tools alone? Over the last decade retail LP organizations have invested significant resources to develop or access diverse datasets ranging from assets (products, employee, facilities) to transactions (POS, inventory, distribution) to exceptions (incidents, exceptions, audits).  However, the current tools and processes that interact with these datasets are both stove-piped and reactive. Take an example of under-ringing. Can exception reporting, electronic article surveillance or video monitoring individually solve that problem, or will it require correlating all of them. Would aggregating them in management dashboards be enough for an effective response or would it require contextualization of information within operational incident or investigation workflows. 

A survey of top 100 retailers shows that the majority monitor their risk based on manually aggregated static dashboards. Shrink is measured once a year, and priorities are determined based on annual comparisons vs. real-time information. Fewer than 20 percent of exceptions are detected vs. reported. Most LP processes, including security, safety, audit, emergency management, incident management, investigations and training, continue to be disjointed, thus leading to a lack of correlation between exceptions and actions. But, can LP afford to use the same tools and processes to handle the most unprecedented risk management crisis in decades?

We believe the only viable answer is for LP to transform itself to a more proactive state. LP needs to drastically shorten the time to detect losses, increase its ability to identify persons, places and things associated with losses and predict future risk in time for maximum prevention. So is such a transformation possible?

Such transformation is already taking place in other sectors. financial service institutions facing exploding identity-theft-driven fraud and anti-terrorism financing have implemented enterprise-wide transaction monitoring and “know your party (customer, vendor, and employee)” solutions. Faced with the changing threat of counterterrorism and cyber warfare, the defense industry has moved to “Network-Centric Warfare.”  It is a transformation from hundreds of siloed sensor data processing systems and databases to “data fusion” and delivering the fused information in the right context directly to the edge (warfighter). LP has the same need: connect the dots, correlate the information into actionable intelligence and deliver it proactively into interactive assessment, prevention and response workflows to field personnel.

The good news is that LP can leverage most of its existing investments in sensors and data. The problem has been that despite the increased availability of data, there is limited ability to connect the dots. What LP needs is new capabilities in data fusion, including (a) automated profiling and detection from transaction monitoring; (b) score-based identification of high-risk entities; and (c) enterprise-wide workflows with correlated information and analytics. Put together these capabilities can provide the interactive command and control framework needed for proactive response.

So how does LP execute this transformation? By implementing what Paul Jones at RILA calls the “Enterprise LP Suite/System” (ELPS). ELPS provides a platform to virtually link disparate data assets and fuses workflow and analytics to deliver composite solutions. First, LP needs to identify the top three areas of perceived losses/risks. Is it fraudulent refunds? Is it under-ringing?  Is it store shortfalls from supply chain irregularities? Is it e-fencing related organized retail crime? Then, implement a “Composite Solution” that combines existing information to

a) Correlate the data sources required to detect or investigate that risk,
b) Score the entities and exceptions involved,
c) Connect the operational investigation and response workflows,
d) Analyze the risk indicators within ad-hoc interactive dashboards, and
e) Automate the discoveries into a proactive detection and monitoring framework

ELPS enabled composite solutions can not only solve the tactical risks cheaper and faster, but also provide a long-term strategic platform for enterprise-wide risk management. By integrating existing applications and databases into a unified context, the platform can support automated detection and drive intelligence-led investigations.

Imagine a world where retail loss prevention can predict and prevent. Customers, employees, products and stores can be continuously scored and monitored for suspicious behavior. Incidents can be investigated with speed and accuracy using the spatial, temporal and link analysis context of similar and related information from all source information across the enterprise including sales, operations, logistics, shrinkage and human resources. Transactions, incidents and alarms across the enterprise can be correlated to predict high-risk events preventing extreme losses before they occur. Key risk indicators can be monitored and correlated on a daily or weekly basis. We look forward to working with the loss prevention community in 2009 to address the immediate challenges and to progress their continuing vision to become fully proactive.