The strength of artificial intelligence (AI) technologies is that they empower users to see both the big picture as well as the granular details: AI can make sense of Big Data and business analytics, so that companies can forecast and plan predictively based on actionable intelligence about consumer behavior, supply streams or staffing activity.
The grocery sector of the retail industry is increasingly tapping the power of AI: Grocery store chains—especially the larger ones—are implementing AI-driven technologies to achieve business objectives, such as improving or expanding customer service, reducing food waste, and ensuring adequate inventory. AI applications for grocery retailers are diverse and solution ecosystem is growing rapidly: One chain even has adopted AI-driven robots to check warehouse inventory, assemble grocery orders and deliver them to customers via self-driving cars. For most, the benefit of artificial intelligence and data-driven approaches is the access to information that enables them to more accurately forecast supply inventory and consumer demand; understand when (and why) to change product pricing; and even which products are “hot” commodities or which ones might go to waste. “Smart Shelves,” another example of AI, use RFID technology to track the goods on the shelf, enabling retailers to monitor how many items are on a shelf and which products were removed and placed in a cart.
Another example of how grocery stores can enhance existing infrastructure with AI technology for data-driven decision making is Video Content Analytics: Many grocery stores have already deployed video surveillance systems for security purposes, such as monitoring various areas of a store or warehouse, and reviewing footage for post-incident investigations. Video content analytics software driven – based on deep learning – enables grocery stores to maximize their existing investments in video surveillance by making video searchable, actionable and quantifiable. Leveraging video content analytics, grocery stores can more effectively prevent inventory loss, improve store layout, streamline the purchasing and stocking of goods and, ultimately, improve the customer in-store experience.
Experts have found that less than 1% of surveillance video is ever seen by human eyes; there is simply not enough time to review all the footage that is collected, and human observations are prone to error. Video intelligence software solves this problem by processing video resources to detect and identify the objects that appear, classify them (people, vehicles, and other items) and then index the metadata to enable precise video search and filtering based on extensive class and attribute combinations, so that footage can be easily and quickly understood and analyzed.
Rapid video review is essential for accelerating investigations and retroactively uncovering gaps in inventory records. Grocery stores can use video content analytics to search and filter video to find suspects and persons of interest after a theft. By identifying past shoplifting offenders, grocery stores can configure real-time alerts for unusual behavior – such as activity in certain areas of the store or stock room at off hours – or even face recognition-triggered alerts when face matches of specific individuals are detected. In other words, better understanding historic data or activity can drive loss prevention strategies moving ahead.
Video data can be used not only to improve physical security, but also to gain valuable business intelligence. By collecting and aggregating video data over time and visualizing the metadata into dashboards and reports, managers can understand trends and activity patterns based on concrete data and make informed, data-driven business decisions. Thus, video analytics transforms video surveillance into a powerful resource for deriving operational intelligence.
For example, video analytics can help grocery retailers track interactions with objects – by displaying “background changes” in the scene as a heatmap – understand the time consumers spend in certain areas of the store, and illustrate the paths customers take when navigating the store space. Based on these heatmaps and navigational data, grocery managers can observe trends of customer activity and optimize store layout based on actionable intelligence. Based on dwell reports and by configuring people counting alerts, managers can also manage queues and crowds predictively and proactively, developing long-term solutions for these challenges as well as data-driven plans for expected traffic peaks and real-time intervention.
This functionality also helps grocery retailers pinpoint obstructions and underutilized spaces, by evaluating navigation trends in-store. Improving traffic flows and optimizing floor planning and display placement are critical for enhancing the overall customer experience in the store. For instance, based on these data points, retailers can optimize strategic staff distribution around the store, based on traffic hotspots where consumers tend to dwell or interact with items. By comparing traffic trends from video data to POS or inventory records, stores can even draw correlations between the dwell time and actual purchasing, for driving future planning and intelligent decision making.
Legacy methods of product purchasing rely too heavily on experience, assumptions and instinct instead of quantifiable data: With access to quantitative data about which products draw the most attention in-store or that are most popular with their customers, grocery retailers can understand which products to purchase and make more intelligent decisions – and orders – moving forward. AI-driven video analytics can complement other inventory and stock management solutions for taking the guesswork out of product purchasing by providing quantifiable and actionable data for guiding decision-making.
It’s important for retail grocery stores to know their audience. Video intelligence software offers granular demographic data, so retailers can understand who is visiting and shopping in-store and derive quantitative intelligence about their customers and their buying habits. By understanding who their shoppers are, grocery managers can cater their efforts towards those audiences, as well as determine which advertising, promotions and products could appeal to other target populations that aren’t as drawn to the store.
Amongst the numerous artificial intelligence technologies offering grocery retailers new ways to improve the customer experience, video content analysis leverages existing video surveillance systems to provide users with quantitative and qualitative data for increasing performance in the individual branch and – ultimately – across bigger chains. With video data providing crucial insights for informing decisions about marketing, merchandising, advertising and layout strategies, grocery stores can optimize their efforts to ensure that shoppers aren’t only shopping, they are buying, and coming back again.
Signup to receive a monthly blog digest.