5 tips for a successful deployment of video analytics in your retail environment

Retail shrink is at an all-time high, accounting for 1.62% of retailers' bottom line. While trying to embrace frictionless shopping methods, you may be unintentionally weakening your security. In this context, video analytics can be a powerful tool to add to your loss prevention strategy.

However, the deployment of these specialized tools comes with its challenges. Here are five crucial elements to consider when evaluating and deploying video analytics to increase success in your retail environment.

 

 

  Define your expectations

“Begin as you mean to go on.” In other words, start by identifying the problem you’re looking to solve. Then, set the right performance expectation, and define metrics for success. Video analytics offer the best insights when deployed as a solution to a problem rather than a solution in search of a problem – the latter is often challenging to evaluate.

Are you looking to speed up investigations using motion detection to filter down to a particular video of interest? Or, do you want to identify suspicious behavior outside of your businesses on off-hours? Taking the time to set these goals and build your deployment expectations will inform how you measure success and, therefore, how you evaluate the different solutions available on the market.

  Know your analytics options

Once you know your needs, finding the appropriate video analytic for the job will help allocate resources appropriately and limit overestimating results. You’ll need to consider each analytic’s intended environment of operation and judge how well it matches your scenario. Using video analytics outside of their intended parameters makes performance unpredictable, often to the detriment of your goals.

Another consideration is the server-based or edge-based nature of the video analytic processing and its effect on your network bandwidth. Both video analytic processing models have their limitations and advantages that relate to your system’s architecture.

• Edge-based analytics reduce bandwidth usage by transmitting only the results of video analytics processed by the camera. However, this requires specialized cameras equipped with the hardware needed to run video analytics.

• Sever-based analytics process video streamed to a server which increases bandwidth usage. However, this allows you to take advantage of the existing video surveillance infrastructure.

For more on video analytics, be sure to read up on how you can use video analytics for more than just security.

  Don’t set it and forget it

After selecting and deploying a video analytic solution, it is crucial to use the metrics you defined to measure performance continuously. Video analytics are not a “set-and-forget” type of technology. High accuracy has traditionally been hard to obtain, especially in open areas with many moving parts and people.

Often, factors in performance issues are video quality as well as lighting and positioning. You’ll find the best results come with continuous adjustments over time. Monitoring performance and working with a trusted vendor will allow you to make the fine-tuning adjustments that make the difference between a fair purchase and a valuable return on your investment.

  Think beyond analytics

Video analytics usually serves as a trigger point in your broader security infrastructure. Your video analytics should link into a centralized, unified system instead of operating in a silo from which extracting data becomes challenging.

The right security system will include event-to-action, alarm management, and map-based monitoring to leverage video analytic data. It becomes even more powerful when paired with a decision management system that can trigger workflows when suspicious activity is detected or correlate multiple events into a single incident trigger.

The power of unified analytics becomes clear when conducting point of sale (POS) investigations. Instead of exporting long lists of POS transactions to correlate them to a video management system (VMS), your centralized platform can refine and streamline everything you need in one place.

  Measure your ROI

When you define the success conditions of your use case before deployment, establishing the return on investment becomes more straightforward. A good example is the people counting video analytics used to prove compliance to occupancy regulations. This use case’s ROI compares the cost of the video analytic solution against hiring staff to count customers and the cost of any violation of occupancy regulations.

Another way to increase ROI is to look for more ways to use your video analytic data. Deploying your security in a unified platform allows you to combine POS data with video recordings and video analytics to pinpoint suspicious transactions. For example, motion detection video analytics can reduce investigation time for fraudulent returns where no customer was present by narrowing down the search to only transactions where a return was completed, but no motion was detected in the area customers would typically occupy.

Thinking beyond the immediate results of video analytics and combining them with other data sources on a unified platform empowers you with a complete view of events in your retail locations.