To help its customers in the food and beverage industry better anticipate demand for new products, Kerry sought powerful AI technology for building a predictive trendspotting solution.
Anticipates trends earlierto help customers go to market at the peak of demand
Provides concepts in 5 daysinstead of the 4 - 6 weeks required previously
Shortens product creation processto < 2 months from ideation to commercialization
Business challenge story
Rapid shifts in food and beverage trends
Offering insight across cultural and geographical barriers, social media is changing the way people live and how they behave as consumers. “Because of social and food ecommerce, there is transparency in the food and beverage landscape that’s never before existed,” says James Sandora, Global Vice President for Digital at Kerry Group. “Now, consumers find food and beverage inspiration from anywhere around the world, and expect the same accessibility in the things they buy.”
Combined with unprecedented access to health and nutritional information online, this shift is creating new challenges for Kerry’s customers in the food and beverage industry. Consumers demand more from their products, and those demands change rapidly.
“At the highest level, the challenge is simply that the food and beverage landscape is evolving and shifting faster than it ever has in the past,” says Sandora. “Our customers have to become extremely nimble in how they address consumer needs.”
Instead of relying solely on traditional research methods, Kerry’s customers needed a quicker way to gauge and keep pace with consumer expectations. Envisioning a cognitive tool for delivering insight into breaking food and beverage trends, the company began exploring AI platforms.
Powerful AI to identify consumer patterns
Evaluating leading AI solutions, Kerry Group chose to build its insight tool with Watson™ technology. “It came down to the service that we got from the IBM team and the precedent IBM had set in understanding how to process natural language and identify trends within the way people speak and write,” says Sandora. “That’s where IBM clearly differentiated itself.”
In a collaborative effort with IBM, Kerry developed Kerry Trendspotter, a predictive platform for forecasting global food and beverage trends. The tool incorporates IBM Watson Natural Language Understanding to crawl through social media content from global consumers and food industry influencers, analyzing it for attributes that include strength of sentiment and emotion.
Combined with online media input provided by IBM Watson Discovery News, the extracted data runs through a machine learning model built using IBM Watson Knowledge Studio and incorporating Kerry Taste & Nutrition’s proprietary intellectual property. The custom model identifies and ranks emerging patterns, predicting ingredients, flavors and foods with the highest probabilities of trending.
“We used the ‘Kerry IQ’ to teach the AI how to think about food and beverage and what we would define as a macro trend,” says Sandora. “At Kerry, we have a huge amount of historical data relating to the ways in which consumers think about, talk about and even interact with the foods and beverages that they eat and drink.”
The entire solution runs on high-performance IBM Cloud infrastructure comprising IBM Cloud Virtual Servers and employing the IBM Cloud Kubernetes Service for workload scaling and security purposes. The environment supports a continuous integration/continuous deployment (CI/CD) pipeline for quick and efficient code changes and updates to the Kerry Trendspotter platform.
Products released at the peak of demand
Together, IBM and Kerry developed a highly effective AI platform that helps food and beverage brands tap into shifting consumer preferences and bring successful offerings to market. “When our customers know what products consumers want and begin development earlier, those products hit the market at the peak of consumer demand and everyone wins,” says Sandora.
The solution has already proven critical for a Kerry customer seeking a distinctive new flavor for a popular snack brand. Instead of the four to six weeks the process typically required, Kerry identified emerging flavors and delivered product concepts in just five days using Kerry Trendspotter. The entire undertaking from ideation to product commercialization took the customer less than two months — a significant acceleration of the previous six- to nine-month timeline.
Combining Watson technology and Kerry’s industry expertise, Kerry Trendspotter is unlike other offerings in the marketplace. “Tech companies are trying to compete in this landscape,” says Sandora. “But we have nearly 50 years of experience in food and beverage to validate trends we’re seeing and deliver a product. It’s a huge differentiator in terms of how our customers are seeing value.”
Based on global customer demand for Kerry Trendspotter, Kerry is working with IBM to deliver translated results in multiple languages for food and beverage producers worldwide. The company credits IBM for an innovative and supportive ongoing effort.
“What I appreciate most about working with IBM is its iterative and collaborative approach to partnership,” says Sandora. “The team we worked with was fantastic in continually improving on ideas so that we could really meet the expectations we had in predicting future trends.”
Kerry (external link), the world’s leading taste and nutrition company, offers solutions that nourish lives all over the world with offices in 32 countries, 150 manufacturing facilities and more than 26,000 employees globally, including over 1,000 food scientists. We bring to the table our strong food heritage, coupled with over 40 years of experience, global insights and market knowledge, culinary and applications expertise, as well as a range of unique solutions that anticipate and address our customers’ needs. For more information, visit Kerry.com (external link).
Take the next step
To learn more about the IBM solutions featured in this story, please contact your IBM representative or IBM Business Partner.