Major food companies are using artificial intelligence, and specifically machine learning, to try to move closer to consumers – and ward off competition from more nimble, digital-first players.

“AI is extremely valuable to food companies because the primary reason for the high rates of failure in new product development is that the consumer and market need is not fully or correctly understood, Kishan Vasani, co-founder and CEO of food innovation intelligence platform Spoonshot AI, which has worked with food companies including Snack Brands Australia and UK manufacturer Winterbotham Darby.

How well do you really know your competitors?

Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.

Company Profile – free sample

Thank you!

Your download email will arrive shortly

Not ready to buy yet? Download a free sample

We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form

By GlobalData
Visit our Privacy Policy for more information about our services, how we may use, process and share your personal data, including information of your rights in respect of your personal data and how you can unsubscribe from future marketing communications. Our services are intended for corporate subscribers and you warrant that the email address submitted is your corporate email address.

Vasani explains AI requires data – “high in volume, velocity, variety, and veracity” – to uncover emerging consumer trends and help companies define and market the more nuanced propositions necessary to get discovered and grow sales.

Spoonshot crawls 28,000 food-related public data sources, including recipes, menus, reviews, social media and news. Machine learning then trains Spoonshot’s algorithms to balance problems like how to weight social data in light of its skew towards younger demographics and informs its Food Brain, a queryable database that assesses future food trends.

Staff at Spoonshot then put the two together to create actionable insights for customers on topics like ingredient substitution or flavour and product extensions. For 2022, Food Brain is predicting nine major trends, including alcohol majors moving into grain-based milks, and more companies releasing lab-grown, fermented products.

Training machine-learning algorithms generally requires access to verifiable data and, in the CPG industry, there are few better sources than e-retailers such as Amazon and, which share their information on consumers in return for products and services. However, over the last two years, regulations like Europe’s General Data Protection Regulation and Apple’s privacy rules have made it more difficult for data brokers to track consumers, changing the nature of competition around access to machine-learning inputs.

“Each retailer is now curating their own proprietary and frequent shopper-type database,” says Ken Harris, managing partner at US-based sales and marketing consultancy Cadent Consulting Group. “Scanned receipts from companies like Fetch Rewards and Numerator, where shoppers scan receipts with their phones and receive points, or money, are another fast-growing source of verified consumer data.”

As a case in point, Walmart has just launched Luminate, a consumer data platform that allows its own teams and suppliers to survey customers about their purchase decisions. The product was put together by Dunnhumby, the same company that assisted Kroger in launching 84.51, the grocer’s own data insights company.

At PepsiCo, AI is helping improve its understanding of consumer preferences, according to Denise Lefebvre, senior vice-president R&D at the US giant’s snacks business.

“We use a suite of AI tools throughout the product ideation process, including an always-on analysis engine that identifies, tracks and predicts consumer trends,” Lefebvre says. “That tool’s emerging ingredients aspect supported the highly successful launch of a seaweed range for crisps brand Off The Eaten Path in the UK, for example.”

At Mondelez International, AI has yet to majorly influence the release of new products but is integral to a digitization drive aiming to put the technology in the hands of product developers. “We work with a range of third parties, from start-ups like [AI consultancy] Fourkind to large established companies like IBM,” says Joe Manton, director for modelling and simulation at Mondelez, says.

The maker of brands like Oreo, Sour Patch Kids, Toblerone and Cadbury has launched a chatbot to interact with geographically dispersed consumers and is using AI to assess their preferences. “Our plan is to have the consumer drive product development with direct input, transforming the way we work,” Manton says.

AI is not required to intuit that consumers are currently clamouring for alternatives to meat and dairy products but Chile-based NotCo is using the technology to make plant-based food that looks, smells, functions, and tastes the same as animal-based counterparts. NotCo’s proprietary AI, known as Giuseppe, analyses thousands of plants in its database to come up with unique combinations that replicate animal-based products.

“It understands and matches products at a molecular level to create plant-based versions of animal-based foods, giving us an advantage due to the speed and accuracy with which we can bring new products to market,” says Karim Pichara, NotCo’s Chief Technology Officer. Giuseppe has helped develop all of NotCo’s products, which include substitutes for milk, meat, burger, cream and mayo. Investors are onboard; in July, NotCo received $235m in Series D investment that gives it a $1.5bn valuation.

Cadent Consulting Group’s Harris suggests every major ice cream company is working with an AI organization to create a perfect [substitute] product. “If non-dairy eventually takes 30% market share, and the total ice cream business only grows at 2-3% a year, if you don’t play you will lose. You have to be in,” Harris says. “It’s not just taste, it has to be the same texture, the same melting point in your mouth, and that is very difficult to achieve.”

A similarly positioned player seeking to take advantage of growing consumer demand for ostensibly more sustainable foods is Motif FoodWorks, which in June tapped $226m in Series B funding to help it solve sensory and nutritional challenges with existing plant-based foods, including meat and dairy alternatives. Before the end of 2021, Motif will debut new technology that provides beef-like umami flavour and aroma, along with muscle protein produced through precision fermentation.

Fermentation specialists using AI to inform their formulations need to be wary about balancing their persona and their products, according to Harris. “Over the next 12 months, you’re going to see a lot of discussion about fermentation, but manufacturers using supercomputers bears no resemblance to fermentation used to make beer; 99% of Vitamin C as an ingredient is fermented fungus – it’s a fine line to avoid frightening people.”

Ultimately, the consumer will judge whether AI-informed alternatives truly match the products they are substituting, but how do you solve the problem that taste is widely acknowledged to be subjective?

Analytical Flavor Systems (AFS), which says it works with CPG companies, uses its Gastrograph AI to model how sensory perception of flavour, aroma, and texture varies across consumer demographics. CEO Jason Cohen uses the example of cinnamon to illustrate how cultural differences impact the way consumers perceive taste: “In the US, cinnamon’s cultural halo is warming winter spices, sugary, sweet flavours. But in Asia and the Middle East its cultural halo is umami flavours – richness.”

AFS’ trains Gastrograph by gathering global consumer preference data the traditional way – central location testing in 25 countries – before compiling a market-segmented breakdown of the flavour composition of individual products and ingredients. This allows Gastrograph to more accurately predict how and why consumers in different markets react to certain foods. For example, the AI determined that the key factor driving consumer preference for ruby chocolate, a variety made using ruby cacao beans, is not red fruit flavours, but citrus and orange.

Cohen raises an interesting difference in approaches to working with AI between partners in the US and China. “In the US we get a lot of validation questions, but in China, everyday interaction with AI is much more prevalent, so the question is: how quickly can you scale this?” Cohen says. “Western companies have catching up to do or faster-moving countries will gain long-term competitive advantage. We may even see foreign competition challenging US companies on home soil.”

Perhaps with this in mind, PepsiCo has just launched two digital hubs in North America and Europe to centralise skills, development and delivery of emerging digital capabilities, such as AI and ML.

“We believe AI is critical to the future of PepsiCo and all CPG companies,” concludes Lefebvre. “Those that excel at it will have a competitive advantage because they will have a much closer connection to the consumer.”

This week, Just Food, along with our GlobalData Media stablemates, has taken a dive into the world of artificial intelligence and how businesses are using the technology. AI is one of the key technologies companies in almost every industry are testing and putting into practice, presenting both opportunities and challenges.

Elsewhere, on Just Food, Nicu Calcea presented a breakdown of the businesses in our sector leading the way in the area.

Over on sister site Shipping Technology, Frankie Youd explored how AI might help tackle the shipping container crisis hitting the global economy.

Other must-read articles include:

Investment Monitor – The 12 companies leading the way in AI

Verdict – Why is AI an obsession for business insiders?

Investment Monitor – AI warfare is coming and businesses need to be prepared