Enhanced Business Intelligence
Start looking at your business in 3-D
- By J.D. Story
- Aug 01, 2011
Almost every industry—including retail, education, hospitality, manufacturing, government, oil and gas, transportation and financial services—is working to increase operational efficiency, manage risk and improve profitability. They all strive for higher performance. In the pursuit of optimum performance and profitability, billions are spent annually on surveillance systems, business intelligence systems, point of sale systems and manufacturing systems. And higher performance is achieved through better management and better management systems.
Traditional Management Strategies
Traditional management strategies include working smarter and harder or hiring more manpower. To improve safety, we hire guards. To improve customer experience and service, we overstaff. To mitigate risk, we hire more lawyers, and to manage better, we hire more managers or manage more efficiently. Hiring manpower is expensive, and it is often difficult to quickly scale and find the right people as an organization grows geographically. By hiring more people and increasing overhead, the risk of becoming less competitive increases. There has to be a more efficient solution.
Many solutions analyze flat or virtual data. However, traditional solutions are independent silos of information and do not communicate or interoperate. Almost none merge the physical world—what we can see, hear and feel—with the flat data from the virtual world. However, a new breed of management system is emerging: video and data management systems, or VDMS for short. These systems combine the physical world with the virtual world, empowering managers and businesses to achieve previously unattainable levels of performance.
In many ways, these new VDMSs are becoming virtual managers, and as such leverage virtual technologies to empower managers to work smarter.
Traditional Management Strategies Using Technology
Using surveillance cameras can help create a safer environment for people and property by establishing a virtual presence. The more cameras and the greater their resolution, the more we are able to see. The problem is that we are drowning in a sea of video, and searching by time, date and location is inefficient. There simply is not enough time to watch all the video we capture. The use of video by the same stakeholders who are responsible for physical security and loss prevention has forced some integration of video management systems, electronic access control systems and point of sale systems. Some video systems have been developed to detect objects, such as people and cars. Many attempt to make searching video more efficient and to tie physical evidence to transactions and events. However, most VMSs are not specifically architected to integrate and correlate non-video sources.
The Potential is Measurable, Significant
In certain retail environments, such as convenience stores, as much as 80 percent of theft is internal. Here, POS data can be indexed to correspond with the date and time of a particular camera, which when combined can quickly identify criminal activity. A major convenience store chain, which has used video and POS with exception-based searching for years, reports that when the system is used daily by store managers, transaction-related shrinkage has been reduced by more than 50 percent.
The key phrase is “when managers use the system on a daily basis.” But as back office duties mount, it is easy for these daily reviews not to be run.
Business intelligence solutions have grown and allow instant access to “flat” data across organizations. These solutions have great capabilities to help analyze information and discover operational relationships to give managers the knowledge they need to take action. They are a powerful tool that, when used properly, can uncover trends and provide valuable insights to make better decisions.
However, many traditional business intelligence systems see the world in two dimensions, lacking visual verification. Traditional systems can tell you that there was a manager override transaction but not whether the manager was present. They can tell you that restaurant revenues are down but not whether the tables are dirty. Can they tell you that the line at Starbucks was too long this morning and people left the line without buying a latte? No. If business intelligence systems could see, feel and hear as you and I do, they would be able to take information from the physical world and add it to the flat data set that traditional systems are designed to collect and analyze.
A New Architecture and Way of Thinking
There is a need for a “virtual manager” that can see and hear what is happening in the physical world, something that can collect, correlate and analyze data as well as apply rules and make intelligent decisions.
To encourage adoption of this concept, a modular approach that will seamlessly use existing technologies would be beneficial. While typical leading multi-site retailers spend a significant amount of money on in-store technology, they do not realize all the potential value. These retailers know less is known about a customer who physically visits a store than about one who shops online.
It is possible to increase performance of physical stores by integrating all the disparate in-store technologies into a single system that can correlate, analyze and clearly communicate performance metrics in real time to in-store staff.
An Example: Retail Shopper Counting
Today, customers show up at stores armed with their smart phones with the latest information and on-demand capabilities. They are accustomed to, and expect, an efficient purchasing experience.
Traffic counters tally the number of people who enter the store. Surveillance cameras provide theft prevention and improve safety. POS systems process transactions and keep track of cash. Each of these solutions has been around for years as an independent solution that lacks integration or correlation with the others. Video is used as an after-the-fact forensic tool, while POS data is flat and provides data that does not tell the whole story. Traffic counters are not accurate and do not effectively filter shopping groups. Traditionally, none of these systems interoperate. Data is not in real time, and different departments typically manage these systems.
Understanding foot traffic in real time now allows retailers to optimize staffing and customer experience and to create employee incentives to turn shoppers into buyers. Simple laser beams are crude, not very accurate and unintelligent, but they do give an idea of retail traffic. Video cameras and thermal cameras can use machine vision to classify objects as people, but their accuracy can be affected by environmental factors such as shadows and temperature changes.
The most accurate, most effective cameras see in 3-D. These cameras have perceptual intelligence. They can distinguish between shadows and real people, and they can filter out children based on height. Because they see in 3-D, these cameras can exclude shopping carts and even define shopping groups. Perceiving the physical world accurately creates a powerful data input that can transform information management systems from seeing in
2-D to seeing in 3-D.
A growing trend is for retailers to compensate employees based on conversion rate. Tying performance to pay is a strong motivating factor and drives productivity. However, accuracy must be high, and employees are increasingly demanding that conversion rates be validated because inconsistent counts affect their compensation. If the video and data from a counting camera could be integrated into a data management system, retailers could easily validate counts. By correlating occupancy data with a POS transaction, such as a manager override or a refund, an intelligent alert and action is created. An e-mail or notification to the loss prevention team that a manager or customer was not present can potentially identify a fraudulent transaction without human intervention.
To accomplish this, it is critical that the next generation of video and data management software solutions allow disparate devices and data sources to be easily integrated and the results to be displayed in an easy-to-understand, unified GUI.
Such key metrics as sales conversion, total sales and average basket size would empower local store management with realtime performance and management data. Corporate headquarters can also receive and aggregate all of the data into an enterprisewide, real-time dashboard.
Because it integrates multiple physical data feeds—such as people counting, occupancy, speed of service and customer tracking— with traditional data feeds—such as access control and POS transactions—in-store technology has an exponential value because it provides for more than just security or loss prevention stakeholders. It is now relevant to merchandising, marketing, human resources and operations staff.
The Future Starts Now
Integration of the physical world with virtual data opens up exciting new possibilities. If we take the retail example above, in the future customers may opt-in to loyalty programs and register smart phones for an enhanced in-store experience. Upon entering the store, the VDMS recognizes the customer’s smart phone preferences and shopping lists. An integrated, purpose-built 3D camera perceives a child in the buying group and adjusts recommendations. The VDMS analyzes and suggests the most efficient route within the store to fulfill the shopping list in the shortest time. As the shopper browses the store, the VDMS correlates data to make recommendations to enhance the shopping experience.
As the system tracks the customer through the store, if it senses a need for assistance, it dispatches the nearest salesperson. The salesperson is updated and armed with the customer’s information, needs and preferences. The shopper’s expectation of receiving a personalized experience is being met with relevant information delivered in real time.
More-intelligent technology makes all of this possible. The pieces exist and are increasing in capability and accuracy, but it starts at the core. It requires a purpose-built business intelligence platform designed to integrate these powerful data inputs with both video and video analytics. It requires a single, unified dashboard that provides your management, operations, merchandising, marketing and IT leaders with an accurate, real-time and recorded view of business performance.
So think big. Consider each of your functional teams. Consider all data that these teams rely on to run their business, both today and beyond. Video-enhanced business intelligence is here, and with the right vision and planning, it will empower your entire organization to deliver significant operational improvement and the best possible customer experience.
This article originally appeared in the August 2011 issue of Network-Centric Security.