38 ON THE TABLE A data-driven recipe: How will artificial intelligence cook up the future of food? FOOD & BEVERAGE ASIA JUNE / JULY <strong>2020</strong>
ON THE TABLE 39 Aroma, flavour and taste are very personal to each individual. But with the advancement of artificial intelligence (AI), these preferences can be translated into useful data to forecast the future trends in foods and beverages. <strong>Food</strong> & <strong>Beverage</strong> <strong>Asia</strong> finds out more from Coralie Garcia-Perrin, Global Senior Strategic Marketing Manager, Sweet and Modulation Taste, Kerry Taste & Nutrition; and Parth Patel, Vice-president, Marketing and Strategic Planning, APMEA, Kerry Taste & Nutrition. Knowing what the future holds is arguably an ability every business would yearn for in order to take leap among its competitors and secure its next success. Unfortunately, the future does hold a degree of uncertainty as no one is able to control the future. That said, the level of uncertainty can, perhaps, be narrowed through the use of technologies like artificial intelligence (AI), which analyses data based on consumers past behaviours and preferences. This set of insightful data will then be key in empowering businesses to better understand market trends, make more accurate decisions, and even forecast what is next on the horizon. In an attempt to forecast future trends in foods and beverages and guide the development of winning consumer-preferred products, Kerry worked with IBM to launch a predictive AI tool, Kerry Trendspotter. Powered by IBM Watson, Kerry Trendspotter is designed to reduce the risk of failures and increase the probability of success, enabling food and beverage companies to use largescale, unstructured data to understand consumer behaviour and anticipate future consumer needs. Parth Patel, Vice-president, Marketing and Strategic Planning, <strong>Asia</strong>-Pacific, the Middle East and Africa (APMEA), outlined two objectives of Kerry Trendspotter. First, it is to increase to probability of success by launching food and beverage products in line with or ahead of trends, and second, to reduce the overall lead time for innovation. He further cited a Nielsen study that revealed the rapid shift of consumer preferences contribute to the failure of 85% of new food and beverage products within two years of launching. He added that failures tend to create inefficiencies in cost, time, and effort for food and beverage companies. “This AI-enabled tool predicts, with a high degree of probability, soon-to-be-popular flavours and ingredients across 60 countries,” Patel explained. “Kerry Trendspotter not only looks at content from social media but also other digital content. Digital media and connectivity are making it easier for influencers to post and share their experiences, and for people to know about each other’s world and desire to experience it.” Hosted on the IBM public cloud, Kerry Trendspotter enables the analysis of food-related social media content by digesting vast amounts of textual data “quickly and securely” through a number of Watson portfolio services such as Watson Natural Language Understanding and Watson Studio models. IBM Cloud infrastructure powers the processing, automatically reading and processing millions of consumer-related posts on social media, extracting food items and cataloguing food-related combinations then used to predict the trending food, ingredient and flavour candidates likely to go mainstream in the near future. By identifying and detecting food trends at their genesis, Kerry’s chefs, food scientists and flavour experts can collaborate with its customers to craft differentiated yet relevant solutions as well as launch first-to-market products with less lead time, capitalising on trends before their competitors via Kerry’s integrated sourcing and development capabilities. He continued, “We are seeing an increased AI adoption of new product innovation (NPI) by food and beverage manufacturers. We have a holistic approach towards delivering insights-led innovation, and Kerry Trendspotter forms a natural fit.”