If you’re in the consumer packaged goods (CPG) business and see a food trend popping up on Instagram or Snapchat, then you are probably already too late to act on it. At that point your company is reacting to what is already happening, rather than what will happen.
This is one of the main drivers of Spoonshot‘s new AI-powered flavor recommendation platform, which the company announced today. This online tool promises to combine data science, machine learning and food science to predict flavor combination trends and provide inspiration for new products for CPG companies.
According to the press release:
“…[Spoonshot’s] platform analyzes billions of diverse data points from over 900 sources in the domain of food. The technology focuses explicitly on long-tail data as stronger signal indicators of future trends are detected in this data. The types of data processed include scientific research, food communities, niche ecommerce portals, media platforms, and data on more than 7.8M CPG products, menus, and recipes.”
We got a chance to try out the new platform, which lets you search through three broad categories: Ingredient Networks, Startups Explorer, and Trend Trek.
Ingredient Networks lets you search ingredients for recommended flavor combinations and pairings. For example, when you search “banana,” it brings up what might be considered unusual recommendations like coffee concentrate and sunflower seed butter. But remember, what Spoonshot wants to do is surface flavors you probably haven’t considered. So bringing up something like chocolate or strawberries would be useless because you already know about them.
As such, each result in the Ingredient Networks search comes back with a novelty score, to tell you how widespread the flavor combination is already, as well as a flavor label to tell you how well those flavors go together based on food chemistry. It also shows you which products already use that flavor combination and any health claims like “gluten free.”
The Startups Explorer flips the search equation and lets you browse through the existing products, menus and recipes in its database to see what flavor combination are already in use.
Though Spoonshot developed this platform for primarily North American use, the Trend Trek feature will eventually allow you to take a trip around the world and pick a country to see what products and flavor combinations are popular elsewhere. Right now it only lets you “visit” Singapore, the UK and the United Arab Emirates.
As part of today’s announcement, Spoonshot said that its AI has created the blueprint for a new product, “an Ashwagandha (Winter Cherry) wellness drink, flavored with Jasmine Flowers.” This combination, the company says, is driven by consumer interest in the power of adaptogens to reduce stress and improve sleep. The company also released a report on how fruit and vegetable peels are an area, err, ripe for new uses and innovation.
Spoonshot has already been working with 40 companies during its closed beta, 15 of which the company says are billion dollar-plus businesses, including Cargill. Spoonshot offers access to its platform through different subscriptions at price points that the company did not disclose.
Using AI to power food recommendations is actually becoming a pretty crowded category. Other players in the space include Analytical Flavor Systems and Tastewise (which also predicted functional foods as an emerging area for CPG companies to focus on).
With all these options, I predict big CPG companies really won’t have an excuse for not getting ahead of the next big food trend.