Forecasting flavor preferences with AI

FEATURED SPEAKER Fi Conference Master Class 2019
Knowing what flavor will work is the conundrum faced by all F&B manufacturers. How can you be sure to develop the right flavor, that will attract consumers for a long period of time and not end as a fad quickly forgotten? Jason Cohen, CEO and Founder of Analytical Flavor Systems, explains that by using Artifical Intelligence, his company can predict and modelise new flavor combinations that will be of long-term interest to consumers

Jason Cohen, CEO and founder of Analytical Flavor Systems

The snacks and confectionery market is very innovative. What is your view on new product developments? Were there any exciting products that caught your attention?
“Snacking has become a global trend due to a variety of factors - particularly the urbanized and fast-paced lifestyle of the millennial generation. I think there are a number of interesting products on the market, but what continues to strike me is the lack of optimization in product portfolios. CPG companies can do a much better job of balancing their product portfolios to target specific consumer cohorts with each product, to reduce in-brand competition. A single-chip brand doesn't need 2+ chili and lime flavors! There is also the issue of unexplored flavor space: there's a whole world of flavors out there that has never been used for snack products.”

“Innovation in this category is certain to continue and AI will play an increasing role in making products competitive.”

Learn more about current challenges and future trends at the Fi Conferences.

Register here

How can AI predict future consumer preferences?
“That's a long and complicated question to answer!”

“New preference acquisition by consumers can be modeled through the current state of the market;
repeat exposure to flavor creates experiences which change individuals’ perceptions, which in turn changes their preferences across multiple dimensions at the same time but at different rates. By modeling the combined effects of multiple exposures and flavor prevalence, we can predict the macro-level changes up to 10 years in advance and the micro-level changes up to 2 years in advance. “

How can AI help in new product development? What can producers find out by working with Gastrograph?
“AI has a range of benefits. As a company, Gastrograph focuses on optimizing flavor, aroma, and texture for specific consumer cohorts and demographics. Better products are better for everyone. 
With AI, CPG companies can: firstly develop successful products with a higher rate of success, secondly develop competitive products that are preferred to the competition and finally develop successful competitive products faster.”

“By helping companies increase their success rate, make every success a larger win, and by helping them win faster - we can act as a multiplier for their innovation strategies.”

How do you gather data about different flavors?
“The hard way! We run our own large-scale data collection globally through our own sensory system (Gastrograph Review). We have a team that profiles the perception and preferences of different countries and regions every week throughout the world. We also have standing panels in NYC and Shanghai for the rapid data collection on products and categories of interest.”

What interesting flavors have you developed within the snack and confectionery category?
“AI can develop flavors and flavor combinations no human formulator would have ever thought of - some of the more interesting ones in markets around the world used our predictions.”

Consumers tend to be more adventurous and seek yet unknown flavors. But how can producers ensure that their product remains relevant for mainstream consumers?
“Some percentage of the population is willing and open to trying new things - but this is a common failure mode as an innovation strategy. Developing fad-products to catch the latest trend is expensive and low ROI. We prefer to work on products that have staying power. Products that consumer love and takes over the competition, so that they can remain on the market for years as a dominant driver.  There will always be new trends and opportunities, but all successful CPG products are built on repeat consumption.”

What are your predictions for the snacks market over the next 3-5 years?
“That it will become more competitive and more important to develop targeted product portfolios to meet the preferences of a more diverse consumer base across the world. We're no longer in a one-size-fits-all market!”

Don’t miss Jason Cohen’s presentation on ‘Artificial intelligence for targeted NPD from conception to consumption’ at the Snack & Confectionary Master Class on 4 December at 16:15.