How FMCG Brands Are Using Next-Gen Flavor Intelligence To Launch More Profitable Products

There’s no shortage of FMCG brands clinging to the old ways of innovation, desperately trying to generate their next big hit from general industry stats, trial and error, and gut feeling. There’s a sense of artistry about crafting new flavors using this triad of efforts — but when you consider that 76% of new FMCG line extensions fail within the first year, it becomes clear that this approach is leading more brands astray than to success.

Years of R&D are lost. Millions of euros down the drain. 2.105B€ per year, to be precise. And yet, smaller and more agile brands are actually gaining brand loyalty and taking market share because they are able to listen to customers and respond rapidly with new line extensions that match trends.

Bigger brands can respond in three ways: acquisitions of agile brands, groundbreaking transformational innovation, or new line extensions. When launched strategically, innovative flavor line extensions are the most cost-effective way to gain brand equity.

Here’s the main roadblock to successful new line extensions: flavor.

84% of consumers say flavor is the top driver for repeat purchases. Not packaging, convenience, or price. Great flavor creates repeat customers, which reduces product failure and increases profitability. (…)

And this explains why small brands are growing. It’s not that their product flavors are objectively tastier, but that they’re better timed and targeted. These brands are light on their feet, connected to consumers, and capable of adapting to the market much faster than larger competitors.

Large FMCG brands need a better, faster way to create on-trend products to stay relevant. It’s not enough to make something delicious. New products must be right for the market.

General stats, long trial and error phases, and gut feeling aren’t enough anymore to accomplish this. That’s why we’re pioneering the next generation of flavor intelligence.

Ever Paired Kiwi With Oyster In A Dish?

The kiwître is a fresh raw oyster balanced over a fruity kiwi tartare served with toasted croutons and a hint of creamy coconut-lime sauce. It’s an unusual pairing, but it has remained the signature dish at L’Air du Temps in Liernu, Belgium for over a decade.

Chef Sang-Hoon Degeimbre came up with the idea, but couldn’t figure out why it worked so well. He called upon Bernard Lahousse, a flavor scientist, to help him understand. Lahousse conducted a scientific aroma analysis of kiwi and oyster and discovered a key commonality: each featured a distinct greenish, sea-like scent that could draw the other flavors together in harmony.

Chefs across Europe began reaching out to Lahousse. Sometimes they wanted to scientifically confirm pairings of their own creations; other times they wanted assistance in the creation of brand new pairings.

Lahousse quickly learned the limitations of modern flavor science, especially when it came to helping brands launch new flavors, so he dreamed up a system that would shatter the current boundaries:

• Generate a precise way to digitize flavor at the molecular level
• Create the world’s largest flavor database
• Track flavor trends to discover up-and-coming flavors

That was 2007. Let’s talk about 2018.

Why We’re Creating The World’s Largest Flavor Database

Traditional NPD in the food and beverage space involves collecting insights from three primary data sources: flavor houses, industry stats, and market trend reports. The goal is to use the sources to understand what flavor to launch next, but even when considered collectively, they offer limited direction because the insights are siloed.

Foodpairing combines these three data sources into a single, dynamic database that uses artificial intelligence to draw insights that are keenly aware of consumer trends and preferences, can comprehend the molecular makeup of ingredients and products, and offers hyper-tailored guidance to brands.

We call this ‘Consumer Flavor Intelligence’ — and it can tell you what variation of “a zero sugar, fat- and gluten free, mocha-choca-vodka latté” to launch next, when, and where.

Let’s look at a few of the ways we collect the data that makes this possible.

• Digitized Food — We separate, identify, and quantify food molecules using our proprietary protocol for gas chromatography mass spectrometry (GC-MS), enabling our AI to analyze the data and discover flavor matches.
• Global Consumer Behavior — CFI scans social media posts, recipe websites, consumer blogs, restaurant menu cards, and other private data from 175,000+ partner chefs and bartenders to discover and structure information about flavor trends, popularity, and predictions.
• A Multidisciplinary Team — We’re not only flavor or data scientists. Our team encompasses a variety of disciplines necessary to make our wholesome approach possible (you should meet Matthias, our astronomer).
• ‘Powered By’ — Our ecommerce API helps online stores recommend more relevant products to consumers and subsequently feeds our database with insights on shopper preferences.
• FlavorID — These data sources interact to create FlavorIDs, personalized flavor profiles for individual consumers that help them discover foods they’ll enjoy (and help brands have greater precision than ever for launching new line extensions).

Combined with the latest research on consumer behavior and trend data, these elements give CFI the complete picture it needs to provide precise, market-aware suggestions.

Finding Matches That Boost Profits For FMCG Brands

Roughly 20% of a flavor experience can be attributed to taste. The other 80% comes from aroma, which is why Lahousse originally theorized — and proved — that ingredients that share key aroma compounds frequently make for great pairings in recipes.

Discovering these matches is simpler than it’s ever been with Foodpairing’s database of 2,500+ ingredients.

Of course, a great match isn’t just about flavor. Pairing dark chocolate and soy sauce, for example, probably wouldn’t work in many markets. To complete the analysis, CFI also takes into account other relevant context data.

• Current And Future Trends — Thanks to our global chef network, we’re able to see what flavor or ingredient is trending now, what’s up-and-coming, and make accurate predictions about what’ll happen next.
• Popularity And Associativity — We give every possible pairing multiple scores that predict how well it will be received by a specific market.

Our machine learning algorithms then suggest a number of pairings that will not only taste great, but that also have high chances of success in the marketplace — and some brands are already experiencing the effectiveness of this approach. Our work with De Kuyper yielded six validated flavor concepts in just two weeks.

“Consumer Flavor Intelligence helps De Kuyper to shorten the development process thanks to fast and clear substantiated flavor directions.”
— Albert De Heer, De Kuyper

We don’t simply point FMCG brands in the right direction. We offer a data-driven shortcut to the right flavors, right markets, and right timing.

The Future Of New Product Development Is Here

Isolated research insights and guessing games gave us high R&D costs, slow time to market, and a 76% new product failure rate.

We’re ready to leave that behind. Are you?

Using machine learning to master flavor science and understand consumers is the way of the future for FMCG brands that want to lower NPD costs, launch more profitable products, and take back brand equity.

Discover more reasons why flavor is the key to creating more profitable products in our free whitepaper: ‘They went for the packaging, they stayed for the flavor’.

Download The Flavor Whitepaper

And don’t forget to follow Foodpairing for more articles and updates (including a deeper look into CFI’s big data capabilities).