COVID-19 is ruthlessly shaping ordinary stores into supermarkets of the future. They must accommodate all the new grocery shopping trends from panic-buying to delivery and pickup options, all with disrupted supply chains. Grocery retailers are responding differently. Some are cutting back on the number of products they sell, some are skipping distributors and contacting product manufacturers directly, and some are giving their brick-and-mortar stores a complete digital overhaul. No matter which line of action your company has selected, AI retail solutions will enhance your operations significantly. Stephen Hawking predicted that artificial intelligence will transform every aspect of our lives. Grocery shops are no exception. The global AI in retail market is expected to reach $24 billion by 2027, rising from $3 billion in 2020 and displaying a year-over-year growth rate of 29.7%. According to Statista, the consumer goods and retail sectors have been using AI throughout 2020 for various purposes:

Statista

If you want to follow the supermarket industry trends and take your existing retail software solutions to the next level with AI, this article will show you some compelling use cases, warn you of possible implementation challenges, and share five tips to help you through deployment.

Physical stores adopting AI to enhance the shopping experience and become more efficient

McKinsey & Company found that even during the pandemic 85% of the US grocery sales still occurred in physical stores. It appears that brick-and-mortar stores are part of the future of grocery shopping. People still like to see, touch, and smell goods. Remember how you feel when you look at colorful packages of sweets or smell freshly baked pastries. Fabio Parasecoli, a professor at NYU’s Department of Nutrition and Food Studies, says that this kind of experience “gives you an idea of choice, abundance, quality, and pleasure.”

Contrary to popular assumptions, physical stores will not be replaced by their online alternatives. But they will rely on AI to tackle the pandemic-induced challenges.

However, this doesn’t mean that grocers are expected to go all in with technology and leave their less technical clients behind. Walmart is an excellent example of a grocery store that does not aim for tech-savvy buyers only but is willing to improve the shopping experience of the average customer. The chain experimented with different technologies to find what matches customer comfort levels. With this learning approach, Walmart discovered that using cameras and real-time data helped them increase overall shelf-stocking efficiency and sales in the meat aisle by 90% and 30%, respectively.

5 ways AI is shaping the future of supermarkets and grocery stores

1. Sparking consumer interest through personalization

Personalized promotions

A few years ago, promotions were advertised in catalogs or through broadcasts. Both options were rather expensive and displayed the same information for all consumers coming to the store. In supermarkets of the future, AI and advanced analytics offer plenty of information on every individual buyer, such as their meal preferences, food allergies, and motives behind their purchases. By employing AI in grocery personalization, retailers gain extensive knowledge of who is walking down their aisles. This approach enables retailers to craft customized promotions to attract buyers and increase sales. Attesting to the capabilities of AI is Gary Hawkins, the CEO of the Center for Advanced Retail and Technology based in Los Angeles: “AI technology can go incredibly deep and continues to learn over time, so it gets better at knowing which items to promote at what price. This will likely lead to different prices for different people, thanks to offers they are sent.” For example, Woolworth in Australia employs supermarket technology of the future to customize its marketing emails considering not only consumers’ taste but also their past shopping behavior. The chain can predict which items every shopper is likely to run out of based on their previous purchases and suggests these items as well. Another example of customization comes from Kroger, an American retail company. When consumers activate the store’s mobile app while inside, sensors detect them and send a personalized selection of items together with the prices via their preferred communication channel (e.g., video, voice, text).

In-store navigation

Effortless, time-saving navigation is a part of supermarkets of the future — especially if we’re talking about spacious retail facilities like the Mall of America, which incorporates over 520 stores and 60 restaurants. It can be daunting to move around such a space. To relieve its customers, the Mall of America deploys location-based AI chatbots, which operate through Facebook or a mobile app and assist customers in finding products and services.

Dynamic pricing

The dynamic pricing concept revolves around using machine learning and artificial intelligence in shopping to determine the best pricing strategy for different products. For this, algorithms analyze data from different sources, such as historical sales, competitor prices, stock levels, and special occasions. One of the tactics of dynamic pricing is cross selling a discounted item (e.g., buns) with a complimentary product (hot dogs) at a full price. This strategy helps reduce food waste by lowering the prices of goods nearing their expiration date.

2. Improving inventory management

Grocery store robots

Supermarkets of the future increasingly adopt robots to tackle inventory-related problems, such as preventing out-of-stock items, incorrect labeling, and pricing. California-based Fellow Robots developed an autonomous retail robot, which can scan your store shelves as high as 2.4 m above the floor daily taking high-definition pictures of products and their prices. Machine learning algorithms examine these pictures, searching for misplaced products and price discrepancies. Store employees can view the results through dedicated dashboards and take immediate action. Additionally, robots can help futuristic grocery store employees cope with the increasing demand for deliveries. Ocado, a UK-based online supermarket, reported a ten times increase in demand since the lockdown in March 2020. The company is using robots to scan the inventory and pick the right items for each delivery.

AI-based forecasting in grocery stores

Another grocery shopping trend is using AI-powered forecasting systems. AI algorithms do not purely depend on the historical data available at stores. They can self-learn and create forecasts even when the data is limited — for example when introducing a new product or testing a new promotion technique. Some older forecasting techniques work on products as clusters and can’t take the local dynamics into account. A good AI model can make predictions about products at a granular level considering local and regional trends.

3. Reducing theft

According to a National Retail Federation survey, retailers lost $62 billion in 2019 due to theft, which is almost $10 billion increase from 2018. AI-powered supermarket technology can detect inappropriate behavior among both buyers and cashiers.

Shoplifting

Incorporating computer vision in shopping can help store security managers identify theft attempts. Sainsbury’s, a UK-based supermarket chain, employed ThirdEye’s concealment detector, which uses computer vision to detect shoppers who take an item and place it in their pocket. The system records suspicious activities and notifies security personnel. Thanks to this technology, Sainsbury’s stopped almost 6,000 theft attempts between September 2019 and March 2020.

Sweethearting

Sweethearting is another form of theft when cashiers fake the act of scanning items at checkout in favor of consumers. Smart store technology can help spot such behavior. Computer vision systems, such as ScanItAll, use ceiling-mounted video cameras to “watch” cashiers and detect sweethearting events, like covering barcodes and stacking items on top of each other. The US-based supermarket chain Piggly Wiggly reported a loss of almost $10,000 per month due to checkout shrinkage at one of its locations. After installing ScanItAll and retraining cashiers, shrinkage costs declined to $1,000.

4. Improving social distancing

As social distancing is here to stay, supermarkets of the future are looking for ways to control the number of people indoors.

Trackers monitoring social distancing

In-store customer tracking is a grocery store technology of the future that can prevent overcrowding inside your store. A German supermarket company Aldi is using AI in shopping through an automated traffic light system, which controls people flow to its stores. When the number of shoppers in a particular location is below a predefined threshold, the light is green, and the door is open for others to come in. After the threshold is reached, the light turns red and the door closes.

Robots replacing human employees in grocery stores

COVID-19 speeds up grocery store robots’ deployment. Big supermarkets see robots as an opportunity to reduce the number of human workers, which makes consumers perceive the store as a safer place. For example, Walmart uses robots to wipe the floor so that buyers will not be surprised by a human worker disregarding social distancing to clean up a stain under consumers’ feet. A common complaint among warehouse workers is that they cannot maintain distancing given their job conditions. AI-enabled supermarkets can incorporate robots in such an environment to reduce the friction among human employees.

5. Improving the checkout process

Statistically, people spend about 60 hours annually waiting in checkout lines.

“In response to the COVID-19 pandemic, the demand for autonomous checkout technology is driving grocers and retailers to innovate and adopt new technologies that keep shoppers safe and streamline checkout.”

- Lindon Gao, Co-founder and CEO of Caper AI, a software company specializing in AI-powered solutions for retail and grocers. Grocery stores equipped with AI have the tools to correctly identify all the items collected by a particular consumer and charge that person’s bank card upon exit without any intervention from store employees. If you don’t want to let go of your staff or deprive your customers of human interaction, you can mirror what Walmart is doing. Cashiers at this AI store assume a new role of a “Host”. The host’s responsibility is to make sure all consumers are treated in their preferred way. If they favor self-checkout, a host will guide them to an open register. If they like the traditional checkout method, a host will pack their items and process the payment.

AI for smart shopping carts

Caper AI produced a shopping cart that uses artificial intelligence to instantly recognize items and measure their weight (if relevant). The cart contains a built-in navigation and product location system, helping customers to traverse the store. It displays available promotions and can compose a list of recommendations. Caper cart also incorporates a credit card reader, enabling consumers to checkout without cashiers. Caper AI has already partnered with the US-based Kroger and Foodcellar & Co, and Canadian Sobeys Inc.

AI-powered cameras for self-checkout

Supermarkets of the future use cameras and sensors to support checkout-free shopping. For example, Amazon Go has 26 cashier-free locations equipped with hundreds of cameras and computer vision algorithms to monitor which products every shopper picks to charge the total amount off their credit card immediately after they leave the store. Additionally, store owners can use AI and video to facilitate grocery checkout. Ireland-based Everseen developed a visual platform where AI watches videos of customers performing self-checkout in real time. The program can identify errors and correct users immediately. For instance, if a customer encounters an item that doesn’t scan properly at the self-checkout AI kiosk, Everseen’s system will register this incident and notify one of the employees to assist the troubled client. Kroger began deploying this system at its stores in 2020.

Smart shelves technology

An intelligent shelf is another application of AI in shopping that facilitates the checkout process. Such a system typically includes computer vision and sensors mounted on shelves. This technology can identify shoppers picking items from a shelf or returning them. It correctly matches buyers and products and charges consumers for their purchase at the end of the shopping journey.

This grocery shopping technology has additional applications. It can display relevant promotions as consumers pass by, monitor stock levels, and notify employees when the store runs out of some items, and detect the presence of ethylene gas, which is released when fresh items start to spoil.

Autonomous self-driving stores

The idea of self-checkout was appealing before the pandemic. A retail startup Wheelys designed a prototype of another grocery technology — a self-driving grocery store with no employees. The store would move autonomously from one location to another. Consumers would enter the shop through a sliding glass door using the Wheelys’ app. They would pick their items, scan them through the app, pack, and exit the store. Their bank card would be charged automatically.

Barriers to adopting AI in grocery shopping

Independent of the industry, implementing AI is a challenge. According to IDC’s 2019 Global AI Survey of 161 retailers, the major barriers to adopting AI in retail include:

• High cost of AI solutions: costs of developing AI systems vary greatly depending on the software type, its intelligence level, the quality and quantity of data that you want your program to process, and the accuracy of algorithmic predictions. To implement an AI application from scratch, you can easily spend $50,000 on a basic version.
• Lack of skilled personnel: The IDC’s authors interpret it in two ways: there is either a general lack of credible AI talent on the labor market, or the retail sector (including grocers) is struggling to attract AI talent.
• Data-related trust issues: Artificial intelligence in shopping applications heavily depends on data, while grocers need to show transparency in data usage to win consumer loyalty. If AI algorithms abuse client data, the relationship will be damaged beyond repair. According to a recent survey by Deloitte, 70% of responding consumers agreed to share their data with grocery stores. This survey demonstrates that the grocery sector comes third after hospitals and governmental institutions if we look at people’s willingness to share their private data. This shows that consumers trust grocers, but there will be dire consequences if this trust is breached.
• Unclear business objectives: AI demands a considerable investment upfront, while it is difficult to find compelling use cases to present to investors. Udai Chilamkurthi, the Chief Architect of Technology Strategy & Architecture at Sainsbury’s confirmed this challenge at the AI Summit London saying: “I haven’t seen a sensible business case that makes sense yet. The technology itself isn’t mature and comes with its own problems.”

How supermarkets of the future can succeed in AI adoption

Despite numerous challenges, the future of AI in grocery shopping looks bright. Here are the five steps that will help you painlessly employ AI at your store and reap the benefits faster.

• Define your vision for the role you expect this technology to play in your organization and the returns on investment. Deloitte recommends adopting the following paradigm: Start small, scale fast, and build iteratively.
• Find the right talent to fill different expert roles. Throwing data scientists in a data lake and expecting them to present new strategies will not yield results. You will need to recruit diverse internal and external talent who can derive, transform, and sustain value in the long term.
• Adapt the organization’s culture to develop the right attitude towards AI. Many employees still view AI as an intimidating black box, which they fear and cannot understand. Make it clear that AI is there mainly to support humans and help them make informed decisions.
• Clean up your data. The amount of data that AI can use is constantly increasing as grocers gain access to information coming in video format, from social media, and via geo-location apps and devices. Despite the data growing in complexity, many retailers overlook the need for a clear data strategy. Brian Kilcourse, Managing Partner at Retail Systems Research emphasizes the importance of data in a successful AI adoption by saying:

"The top challenge is dirty data (it's the elephant in the room). Models are only as good as the data that creates them. One recent study estimated that over 80% of the effort of implementing AI relates to data cleansing.”

• Gain an ecosystem advantage. Artificial intelligence capabilities for grocery stores are evolving fast, and working with an ecosystem partner will give you access to the right team to get the job done. For example, a grocery retailer can join forces with predictive analytics and demand-sensing company.

Final thoughts

Supermarkets of the future are not necessarily online stores. Implementing AI in grocery shopping enables you to overcome all the pandemic-related hurdles while still letting your customers enjoy the sight, smell, and feel of products. However, incorporating AI is not a one-time task. It is a long process that requires changes of a store’s internal processes and culture. As Sanjeev Sularia, the CEO of Intelligence Node, puts it: “Retail organizations often get deterred by the costs of building the infrastructure and data processing capabilities needed for AI adoption. However, flexible businesses have successfully integrated AI across all business functions and upskilled their people to efficiently reorient to a data-driven mindset without trying to build everything from scratch.”

Thinking to enhance your grocery store with AI capabilities? Feel free to contact ITRex retail and AI experts, and they will make sure your supermarket is well-equipped to face challenges of the future.