Introduction

Video analytics has become an integral tool allowing to control the business of any retail location. The use of video analytics systems helps to meet a wide range of issues: from counting visitors to identifying unwanted customers.

In this article, you will find how we can increase the income of any sales point with application of the modern video analytics solutions. I will also share some examples of businesses that have successfully implemented and currently use these systems.

1. Store traffic analysis

One of the most widespread functions of video analytics in retail is the analysis of store traffic. Such analysis remains extremely effective since it can solve several problems at once.

Visitors counting

Customer counting function can be used to track the number of people in the store and analyze the data needed to estimate the effectiveness of the trade and product layout. The visitor counting function helps to streamline daily work to improve service quality. Used with cameras connected to an IP network, it allows you to access the data you need anytime, anywhere.

What are the benefits of the visitor counting feature?

• Determination of the time of greater workload of the outlet;
• Store traffic analysis by day and time, as well as comparison of traffic with other retail points;
• Assessment of the merchandising and promotions efficiency based on the attendance model;
• Store personnel distribution.

For example, iVideon company installs a smart video counter named Ivideon Counter, which can analyze the traffic of the sales point and helps to plan and evaluate the effectiveness of marketing campaigns. Since the counter is integrated with the cloud service, retailers can receive statistics and watch videos both online and from the archive on any device that is connected to the Internet. (4)

Face recognition

Face recognition helps to improve the security of the sales point, simplify consumer interaction, and conduct marketing analytics. At the entrance to the store, it’s necessary to install cameras that can independently recognize faces and add new visitors to the store’s customer base.

In the system, one can create “white” and “black” lists of customers, and if the camera recognizes a VIP-client or an unwanted visitor, the system will automatically inform the store employee.

Based on customer analysis, managers receive complete data on the composition of their store audience, time visits, and any other necessary data to track performance indicators.

The system is also able to recognize the gender and age of people visiting the store. Age is determined at specific intervals (for example, up to 20 years, from 20 to 30, from 30 to 40, etc.). Modern cameras are capable of recognizing the gender and age of the visitor with an accuracy of 90–95%.

The collected data gives retailers an understanding of how many people visit the store, who are new customers, and who are regular ones, and at what time the store is fully loaded. Besides, the system will show the percentage of the audience by gender and the age composition of visitors.

All this together helps to organize the efficient operation of the sales point to improve the quality of customer service and the company’s income.

Already today, video analytics offers to use face recognition as a tool to increase loyalty. The use of new technology in retail is identifying VIP-customers. With this system, customers no longer need to constantly present a discount card, since the discount system will be triggered based on their biometric data.

Queues analysis

Another function of video analytics helps to control queues. The system informs when the number of people in the queue exceeds the established threshold. Such a system not only analyzes the queue length but also solve a number of extremely important related problems:

• Provide data on the average time of a visitor in the queue. This data will be used by the marketing services of the sales point.
• Determine the number of visitors who left the store without waiting for their turn of service. This indicator will provide information on the number of dissatisfied customers.

Solving these problems and collecting statistical data are very important for optimizing the operation of a retail outlet. There are various illustrations of successful implementation of such a system. For example, on December 24, 2019, X5 Retail Group announced that the company had patented the invention of an automatic system for analyzing delays at cash desks. (9)

The system consists of a program and three components: information sources (cash registers and POS terminals), an integration server and an analytics server.

The program collects data from information sources of a point of sale, then transmits and processes the collected information on servers, determines the sequence, logic and connection between automatic and manual operations.

After all, the system independently draws conclusions on improving processes and the reasons for delays at cash desks. The analysis is carried out in real time and allows managers to know in advance and prevent 98% of problems and failures.

Since 2018, this solution has undergone a phased implementation and improvement. As of December 2019, the system works in all 15,752 stores of X5 chains. It’s constantly being refined and improved through the introduction of updated functionality.

Heat maps

Another effective function of video analytics in retail is the heat map. This feature can provide statistics on the number of visitors who have stopped at certain shelves or products. If we add smart cameras with the Eye Tracking function, we will get the most complete information about the interests of certain groups of customers.

Further, this information will serve as the basis for more effective marketing of goods and more precise planning of retail space. The store manager may decide to organize easier passageways to low-traffic areas of the shop to better sell the goods located there. (1)

2. Layout of goods

Artificial intelligence integrated into special surveillance cameras analyzes the occupancy of shelves with goods and helps to monitor the quality of merchandising. Artificial intelligence is promptly notifying about the problems arising in the display of products that an ordinary employee may not notice (e.g. empty shelves).

Sberbank, the Azbuka Vkusa retail chain and the Visa international payment system unleash the power of video analytics in retail. On May 20, 2020 the companies opened a store without cash desks and sellers, however, equipped with a computer vision system. (7)

First of all, the user needs to download a mobile app developed specifically for this service. While entering the store, the visitor scans the QR-code from the app and starts shopping. The buyer takes the goods from the shelf and just leaves the store.

The money from the linked card will be debited automatically, and the client will receive a push notification to the phone about the transaction and a receipt from the store by email.

The video surveillance system constantly monitors and analyzes the quantity and range of goods in store. The system accurately collects a virtual shopping cart with products that the customer actually chose. If suddenly a person changes their mind and returns the product to the shelf, then the system deletes this product from the virtual cart.

Thus, queues, cashiers or self-checkout counters are no longer part of the shopping process, which saves visitors’ time and reduces staff costs for the store.

3. Staff control

To maximize the profit, you need not only to know the client but also organize the work of the staff. Video analytics helps the retail industry in this matter too.

Working time control

Video-analytic complex CasRetail was created specifically for economic and marketing analysis and control. This system generates a detailed report, which, together with video data, helps to describe the behavior of visitors and employees of the salespoint. The system can also analyze the staff activity, which will allow the owner to get measurable KPIs for each employee and see the productivity of their work.

The video surveillance system from the iVideon company also allows the managers to monitor the observance of the working day, control the quality of customer service, ensure the safety of property, and quickly resolve any disputes.

How to create the employee’s time tracking report? The entire territory is divided into zones where the system analyzes the time of presence of individual workers.

Cameras with modules are installed at the entrance and exit from the zone. These cameras will identify the employee by the face and record their arrival / departure time. In the working areas themselves, cameras will track and record violations of the temporary working regime.

When the managers decide to check the employees, they will be able to set up additional filters and view a report on the working hours of a specific employee or a specific work area, or the manager can set the time interval of a zone in the parameters and view a report on all persons recognized at that time.

So, it’s already possible to control the working hours of employees and monitor the fulfillment of work obligations on time and without overtime. (10)

Work transparency

Video analytics bundled with smart cameras can be used to reduce store financial losses due to staff misconduct. One of the possible solutions is to track the process of selling goods using cash analytics and a video surveillance system.

The seller can use a very simple and at first glance innocent scheme of deception. Often there are employees who register a store discount card for themselves or their relative / friend.

Using their loyalty card with every purchase during a day, the employee provides an undeserved discount to all customers and collects bonus points for themselves. At the same time, we can note that the employee collects bonus points for themselves, and the discount reduces the profit of the store owner. That means that such an employee steals from their employer.

Fixing POS transactions together with a video surveillance system is precisely designed to prevent such fraud schemes. The algorithm for tracking suspicious discounts is one of many that allows managers to draw the controller’s attention to the cashier, in whose shift the same discount card was used more than once.

If you have a cloud server that records data on all such transactions, you can track the use of the same loyalty card not only within one store but even across the entire network. (6)

The Russian company that operates the retail chain, Magnit, is implementing video analytics throughout Russia to rank cashiers. On June 2, 2020, Magnit began to implement video analytics tools in all of its retail outlets. These solutions are designed to improve employee performance and services in stores.

The collected data will help to make an accurate rating of cashiers, which, in turn, will help to predict their work, based on the speed of customer service, the number of too long queues and other indicators.

This will allow the store management to more accurately plan shifts, distribute the workload, make a decision on the need for additional staff training, personnel rotation or the installation of new cash registers to achieve maximum efficiency of the outlet. (8)

4. Control of marketing campaigns

Video analytics technologies give an understanding of which ad is the most effective. It helps to correctly plan the marketing costs and increase conversion.

There are intelligent video analytics services (for example, Seemetrix) for retail that determine the gender and age of a person, and even the time of viewing ads through cameras installed on digital displays in stores. The service collects the aforementioned anonymous audience data and analyzes the effectiveness of advertising, providing a visual report for the user. (5)

The system catches every eye directed at the screen, after which the collected data is sent to the server in real-time. Statistics can be viewed online at any time in your personal account.

These types of video analytics applications allow us to measure the flow of visitors and conduct demographic analysis, as well as correctly place products by customer behavior and show only targeted ads, identifying the products that are most interesting to the buyer.

Facial recognition systems help to improve the efficiency of the following business processes:

• Optimization of advertising budgets, assessment of the attractiveness of certain shopping centers or individual stores;
• Tracking the routes of visitors to arrange the goods in the most profitable way;
• Segmentation of people by gender and age will help to determine the popularity of a particular department of the store among different age and gender groups.

Besides the proper planning of your ad campaign budget through instant audience analysis, video analytics can become a means of additional income.

The market of video analytics

With all the advantages of using video analytics in retail, many companies face certain difficulties in its implementation. The most common factors holding back the growth of video analytics in retail are the need to install a large number of video cameras to cover the entire area of the store.

For example, to conduct an effective analysis of the interaction of a buyer with a product, it’s necessary to fix the camera no higher than the eye level of the client. It’s often very difficult to find a suitable place in a store for the correct placement of all video analytics cameras.

The costs of implementing video analytics can be prohibitive. For example, some shopping centers offer too expensive Internet, while for video analytics to work correctly, one needs to have a constant connection. Also, for video analysis in sales areas with high traffic, it’s necessary to take into account the cost of expensive high-quality equipment.

According to Tractica, despite some setbacks, retail is the largest sector using video analytics services and systems with a volume of about $800 million in 2019 and forecasts of growth to $1.1 billion in 2022 with an average annual growth rate of about 20%. (2)

Conclusion

Video analytics helps retailers to gain valuable insights into the effectiveness of marketing campaigns, such as store traffic, customer demographics, and emotional profiles. Expose KPIs to employees, optimize the display of goods and processes at the sales point.

By the constant analysis of the outlet, the system may count visitors, control the waiting time of the buyer in the queue, and the time of service at the checkout. The integration of the face recognition system increases the security and customer focus of sales points by identifying VIP-visitors or suspicious individuals. And a store heat map makes merchandisers work easier and more efficient.

Video analytics also helps to monitor the work of staff. The system not only tracks the employee’s working time but also identifies the possible strengths and weaknesses of the seller. After receiving a detailed report, managers will be able to make the right decisions to organize the work of the retail point.

Video analytics can also improve the effectiveness of advertising campaigns. The system reports information to make a more targeted ad, track its performance and customer acquisition, as well as its conversion into real sales.