Video Analytics is now the “go-to” retail tech to understand various aspects of retail operations.  It would appear almost all large retailers have experimented using this strategy at some point.

Given the power of video analytics as a technology and the rate at which it has progressed, one is not surprised. Supporting technologies like big data, data analytics, cloud services, software as a service (SaaS), computer vision, machine learning and artificial intelligence have all contributed to this technology getting a mega boost.

That is the bright side of things.  The question is, is this powerful technology being put to the right use?

Our analysis of the market reveals that this is where a huge gap is emerging. Many users are getting sucked in by the sex appeal and hype of real-time data.  They are investing huge capital into installing and wiring up proprietary and expensive cameras and are getting stuck with either very little data or highly unreliable and inaccurate data.

Tech providers, touting “state-of-the-art” or “cutting edge” technologies and algorithms, rely on freely available algorithms that detect humans and identify basic demographics like gender and age group to flood the data with indistinguishable “shoppers”. These numbers frequently include staff who “enter” multiple times during the day and often miss out on individual counts when whole groups or families enter at the same time.

Another ploy that retailers fall for is the camera getting “spoilt” or disconnected mysteriously when inconvenient for store managers.

More importantly, store managers have little clue on what to do with all this “real-time” data and it is used only occasionally to see the trend.

On the one hand, vendors are just skimming the tip of the data iceberg of data that VA can provide and flooding their clients with less valuable & scantily used data.  The fact is that the depth of data Video Analytics is capable of, is really not best used on a real-time basis.  Considering the camera sees practically everything that happens in the store, a meaningful and comprehensive analysis will give highly valuable information about:

  • Shopper profiles
  • Shopper Behaviors
  • Staff Attendance
  • Staff performance
  • Planogram compliance
  • Stock appeal
  • Stock turnover
  • Space utilization – and much more.

The key point is that such data is practically useless on a real-time basis.  It’s best used for trend and comparative analysis by experienced managers or data experts if not scientists. Essentially the extensive data that VA can provide needs to be “morselized” for easy digestion and understood in context.

The experience of other sectors that use VA extensively bears this out.

Take Sports, for instance.  Cricketers do not use VA for real-time analytics.  Rather it is used by coaches who study how many times Virat Kohli got out to an off-side ball and what can be done about it.

Read more about this powerful technology and how best to use it at Retail Fails -Video Analytics: Intervention vs Strategy

Or click on: https://tinyurl.com/How-To-Use-Video-Analytics 

#ideotics #videoanalytics #CCTVanlytics #shopperprofiles