We all know how important it is that we understand our customers really well.
If you’re in retail, with a business of some size, you’re already crunching the numbers, analysing buying patterns, figuring out which customer segments visit the store at what times, what they tend to buy and, if they buy item X, how likely it is that they are interested in item Y as well.
Retailers have data warehouses and advanced computer software to detect patterns and find new trends in customer behavior. Customer analytics in retail is already standard hygiene.
At the same time, everything appears to be in flux.
In the coming years, digital and physical shopping environments will become more diversified as well as connected - or augmented, if you will. With so much information coming to them and so many alternatives at customers’ disposal, their expectations and behavior are constantly changing.
Success comes from knowing your audience and tailoring your message to match their interests. Instant gratification has become the norm. Consumers expect that, from our first interaction, we know everything about them. That we meet their every desire with just a few swipes of their finger.
New technology to the rescue
Now, a new set of emerging technologies is about to make customer analytics in retail a lot more powerful. And since it’s such a mature and competitive sector already, only those who adopt these new methods are likely to survive and thrive.
The internet and the lowering cost of hard- and software have brought about a new abundance of data. When merging data streams of various sorts, artificial intelligence algorithms can detect patterns in those data that humans might never think of. Machines can be taught to recommend adjustments in the retail environment to make buyers enjoy their shopping experience better, and indeed spend more money.
Imagine being able to analyse which customer segment is the happiest with a certain product offering or shop interior, and what type of customer is most likely to visit your shop at a specific time of day. If you knew that in advance, you could move your shelves and furniture around, adjust the lighting, and prepare special offers to seduce the right audience at the right time.
The data boom is just beginning
With Advian’s video-assisted Edge AI solution, this is becoming a reality. Our cameras and software can reveal which customer segments frequent your shop at what times, which routes they take, what they browse, where they pause, what they buy and - importantly - what mood they are in.
Yes, you read that right. Our cameras can “see” how happy people are.
None of this would be possible without enough data, artificial intelligence, and computing power. The latter is achieved by moving the software close to where the data are collected. In other words, computing at the edges of the digital network.
Sending all the video data from multiple cameras to a cloud or centralised server, running it through software and then sending the recommendations back to the shop would be too slow and prohibitively expensive. There is just too much data, and we’re only at the beginning of the data boom. Centralised computers could not handle all the processing, and transferring the data back and forth would require more bandwidth than we will ever have.
Plenty of use cases
The solution is edge computing. With our Edge AI, we collect the video data, interpret them close to the camera, and only send the relevant outcome, the conclusion, for further processing or action.
The software on the camera only needs to tell us if the person on the video is a man or a woman, what age bracket s/he is in, and if they seem happy. Most other information, including any information that could identify the person, is disregarded.
Now that we know that a customer is happy and belongs to a certain segment, perhaps our machine has learned that he or she might be susceptible to a cross-sell or an up-sell. The AI could tell us how to keep them happy and keep them buying.
Or if a customer seems unhappy, our AI might know how previously unhappy customers have responded to different offers. It might tell us which course of action is most likely to make this customer happy, too.
But there are plenty of other use cases as well. Think for example about analysing which parts of a department store are under-occupied or predicting when queues are likely to form at checkouts. With such information we can optimise the interior and allocate shop personnel to where they can have the biggest impact.
Start your Edge AI journey today
Due to the COVID-19 Coronavirus pandemic, social distancing has become a serious concern. Advian’s video-assisted Edge AI solution can provide better insights as to when people in certain risk groups, such as elderly people, come to the shopping mall. Our system can detect to what extent people keep safety distances.
In one commercial solution we’ve installed our video-assisted Edge AI solution with eight cameras at different entrances of a large Finnish shopping mall. Thousands of people will be filmed and their anonymised data fused for analysis.
Advian is a strategic partner in this project due to our expertise in Edge AI and hardware integration for this particular use case. We know what kind of algorithms and cameras are needed and how to connect analysed data to retailers’ business processes.
Edge AI data analytics is about to take a huge leap. For those interested in finding out more we’ve put together an eBook, ‘How to Get Started with Video-Assisted Edge AI for Retail Customer Analytics - and Why You Should Care’.
💡 How video-assisted Artificial Intelligence at the edges of digital networks is about to revolutionize retail customer analytics
💡 Which Edge AI use cases may benefit your organisation
💡 How to start your Edge AI journey by taking your first steps today.
Wonder what your business could achieve
with video-assisted Edge AI?
Book a meeting with our expert below 👇