AI now creates your meal plans at this shopping centre. Is it worth it?
A new era of food shopping has entered the largest shopping centre in the Southern Hemisphere. Find out just how easy it will make your next grocery run.
Artificial intelligence is now here to help you shop better – and choose your groceries faster.
Today, thousands have flocked to Chadstone shopping centre in Melbourne’s south-east to see the newly opened Market Pavilion, a fresh food and dining precinct topped off with a Food Concierge service. An integral part of this premium offering is its AI recipe generator which guides customers through their shopping. Chadstone claims it is the first shopping centre in Australia to use generative AI to enhance customer experience.
“We’re certainly not aware of any other AI tools that have the same range of inputs as our recipe generator,” says Chadstone centre manager Daniel Boyle, who is at the helm of the Market Pavilion project. “I haven’t seen anything that provides an addition of where you might buy special ingredients or how to find value options if you’re concerned about the cost of living.”
While AI is the buzzword on seemingly everybody’s lips at the moment, the uptake of generative technology has been slow in Australia, particularly in the food and beverage sector.
Professor of AI at the University of Queensland (UQ) Shane Culpepper has seen more hesitation in consumers when it comes to the uptake of AI technology in Australia. “In the food and beverage industry, Australia has not yet heavily invested in AI in the way many other countries, like Japan, have.”
“In Japan, you’ll see restaurants or supermarkets that are completely automated. Customers can order food based on preferences and the AI tool will give you recommendations on what to buy. I’ve never seen anything like that in Australia,” he says.
The AI recipe generator sits within Chadstone’s growing premium service offering known as the Food Concierge. Customers can use other elements like the parcel storage service – where goods are stored at a cooled temperature while customers continue to shop – and kerbside delivery.
Boyle and the Chadstone team saw an untapped space in the market for customers with fast-paced lives. AI was part of the solution to make the shopping experience more efficient for shoppers, particularly families with more mouths to feed.
“Chadstone has a really fabulous array of premium services but it was about four years ago, when we really started to get into the nuts and bolts of delivering this beautiful market, we asked ourselves: what will those value-add services look like for customers who are purely focused on food?”
The precinct’s generative AI tool directs shoppers towards recipes created from their preferences and where in the precinct they will be able to find the necessary ingredients. This innovation comes off the back of supermarket giants Coles and Woolworths trialling AI-powered smart trolleys to give customers real-time scanning and tracking updates while shopping in-store.
“When I reflect on my own family and friends, we Melburnians have a real epicurean spirit and love to explore food, but we’re time poor – we get stuck in the cycle of making beef and three veg or our favourite lasagne,” says Boyle.
“The tool was essentially brought to life to make meal planning simpler because we wanted customers to save time and energy, and reduce waste.”
Prompts include the number of recipes needed, the range of ingredients, simplicity of the recipe, number of serves desired, and dietaries. There are also free text boxes where customers may enter ingredients they want to include or exclude entirely from their recipes. The recipes generated focus on ingredients that are both seasonal and can easily be found at the Market Pavilion – whether from luxury goods stores or more affordable fresh food retailers.
The Market Pavilion has an investment of $480 million and buy-in from over 50 food and beverage retailers. The precinct is home to Melbourne stalwarts like dessert brand Brunetti and cheesemonger That’s Amore. Sydney-based butcher Vic’s makes its first brick-and-mortar foray into the Melbourne retail market after supplying quality cuts to fine diners such as Quay, Bennelong and Firedoor.
Janet McColl-Kennedy, a professor of marketing at the University of Queensland’s business school and lead at Australia’s Food and Beverage Accelerator, says that customers’ food choices can be greatly enhanced through AI because “it has the potential to satisfy cognitive needs and reduce decision inertia”.
Decision inertia affects customers when there are just so many choices to be made relating to food. Is there too much gluten? How many saturated fats are in this? What are the total calories?
“When there are multiple decisions to be made in a short amount of time at the supermarket, it can be so overwhelming for a consumer.”
“Filtering methods provided by AI in the form of recommendations can reduce the cognitive burden on consumers,” she says, “it is important that they’re given the choice to set their… tastes and preferences among different food groups.”
Dietitian and senior lecturer at the University of Sydney Juliana Chen welcomes AI as a tool for suggesting recipes and meal plans based on dietary preferences, health goals, and nutritional needs. However, research is still emerging as to whether AI can support the improvement of certain nutrition outcomes.
“While AI can provide recommendations, consumers must understand their own dietary needs and actively engage in their health. Over-reliance on AI may lead to generic, one-size-fits-all advice,” says Chen. “It can support, but it shouldn’t replace individual responsibility in making informed food choices with the support of nutrition experts.”
While not claiming to be optimised for health outcomes, The Market Pavilion’s AI tool has bespoke elements that can make it a useful guide for customers. The recipe generator also caters to dietary requirements so customers can input preferences for low FODMAP, gluten free, low sodium or vegan meals.
“Entries can be as bespoke as the customer likes,” says Boyle. “They might say, ‘create me a recipe where barramundi, chilli and mushroom are used in the base, but I don’t want onion or additional salt.’”
Boyle explains there are two touchpoints for customers to interact with the AI component of the broader Concierge Service. The first is a tablet self-service system for customers to input their food preferences based on a series of prompts. The final recipes generated from this survey are accessible through a PDF format which can then be scanned from a QR code.
The second way to use the tool is directly through the team behind the physical concierge desk who will walk customers through the tool, curate the recipes and provide a printed copy.
Chadstone engaged AI Consulting to create the generative AI tool. The recipe generator runs on Open AI’s o1 and o3 models. Although it is modelled off similar programs like ChatGPT, AI Consulting’s managing partner and AI & data strategist Rad Anandakumar says the Chadstone tool is different because of the centre-specific approach.
“You could go to ChatGPT to ask for a recipe but the answers are usually pretty poor. We’re focused on really accurate and valuable answers that align to the person’s preferences,” says Anandakumar.
Culpepper notes that academics are currently exploring the trade-off between personalisation and privacy when it comes to using AI.
With respect to the AI tool being implemented at Chadstone, he says, “I don’t believe there is a huge risk when it comes to consumer data. If people are treated anonymously, it can certainly be done safely.”
Boyle is cognisant of the concerns regarding data collection and AI. “This is a free tool for our customers to use only within the Market Pavilion. There’s no barriers – no names or email addresses. We know some customers are uneasy about signing up to new technology.”
The tool was refined over the course of six months. While the centre trials the generative AI tool, there is currently one device servicing customers. Boyle says there is capacity to expand to a second tablet if there is demand.
Boyle and the Chadstone team will continue to work with AI Consulting group to develop stages two and three of their generative AI project.
What’s next in tech for Boyle and the Chadstone team?
“Rad and I have enjoyed a couple of conversations recently about how we could use AI technology in the future. This may be the only AI tool at the moment, but the boundaries are really limitless.”
To join the conversation, please log in. Don't have an account? Register
Join the conversation, you are commenting as Logout