I recently caught up with Natalia Hernandez, Data Scientist at Foodpairing, to highlight how her company’s data scientists mine public online data, which gives general trend insights to use consumer intelligence and molecular analysis of ingredients to forecast the next big flavors in the food industry. Natalia Hernandez is a Data Scientist, passionate about making sense of data by building algorithms that provide actionable insights as well as explaining complex technical concepts to people with non-technical background.
insideBIGDATA: Please give us an overview of Foodpairing and the role of data collection and analysis in Consumer Flavor Intelligence
Natalia Hernandez: Foodpairing is a fast-growing company from Bruges, Belgium. We forecast the next big trend in consumer flavor preference through a multi-dimensional approach to market research.
We offer brands help in solving the question “What flavor should we develop next?” Of course, when companies try to answer this question, they begin spending too much time in the fuzzy front end. When talking about product development, the fuzzy front end is the time between idea conception of a new product flavor and the beginning of its development. At this stage in flavor line development surveys are ineffective and expensive and trend reports are too general to provide a business with actionable insights. Companies are left to rely on intuition – a costly risk.
We call Foodpairing’s approach to building new flavor lines Consumer Flavor Intelligence (CFI) through a process that shortens the fuzzy front end of the product development process. Consumer behavior data is one of the elements at the core of CFI because of the massive insights it gives into what consumers are looking for in new flavors and tastes.
insideBIGDATA: Your approach to market research is probably one that would interest our readers the most – how does the Foodpairing team approach the FMCG industry differently than others?
Natalia Hernandez: We noticed a few challenges for companies looking to develop new flavor lines and set out to resolves them.
- First, consumers are no longer exclusively loyal to brands – the millennial flavor palate is evolving at a rapid pace and new products are constantly hitting the shelves in order to keep up.
- Second, 76% of new product line extensions fail within a year – we attribute this to a lack of precision in the fuzzy front end of market
- Third, a new generation of agile companies have entered the markets. Local craft producers, for example, create strong brands with high quality, on-trend products. Their development cycle is shorter than multinational companies, creating a shorter time to market.
Together, the above factors have created a generally risky environment for companies to try out new flavor lines. To address these numerous challenges, we created a solution that surveyed the market from a more holistic perspective.
Our data scientists mine public online data, which gives us general trend insights. We then complete an analysis using our molecular approach to food items. Here, we analyze aromas and measure their compatibility with other flavors. We also complete a brand analysis, which uses qualitative methods to describe the brand itself, including its visual identity and tone of voice, as well as the particular product’s attributes, such as texture and taste. We then can score flavor options against our analysis of the brand identity.
insideBIGDATA: What are the benefits of this method? How does Foodpairing ensure that its findings and proposals will be a success when they hit the market?
Natalia Hernandez: The central benefit to a market research approach that embraces science, consumer data, chef inspiration is that it reduces the cost of trial and error for R&D teams.
The current toolbox in the fuzzy front end isn’t sufficient anymore. Marketing managers, development chefs, and R&D managers lack actionable market and consumer insights. Nowadays, research techniques work great in evaluation, but not during the idea generation phase. When data is at the core of ideation, stakeholders from marketing to R&D can jointly take part in decision making, align on corporate goals and help increase the chances of a successful launch.
For us, it’s about ensuring that the right flavor comes onto the right market at the right moment. This is key to make a new flavor line successful. The market research we provide is tightly linked to consumer preference at that moment and into the future.
insideBIGDATA: Practically, how does this work?
Natalia Hernandez: We provide brands with three tangible reports that are a result of our consumer behavior and scientific research.
We begin by integrating scientific data at the molecular level – our food scientists have developed a way to analyze the aroma of ingredients at the molecular level. We can then say which ingredients would go well together based on this molecular analysis.
Next, we use public consumer data from search engines and social media, trends detected via blogs & recipes sites, and chef data from our own website to gain insights into what consumers are looking to try and thought leaders are inspired to create.
Last but not least, we suggest flavors that match the brand attributes and the target audience.
Market research in the FMCG industry can be limited by its reliance on antiquated methods for arriving at a novel product line extension that put too much responsibility in the hands of consumers who probably don’t know what they want “next.” Our research and subsequent recommendation of possible flavor options is thus a composite of science, brand identity, and consumer behavior.
insideBIGDATA: What’s the next frontier for Foodpairing?
Natalia Hernandez: We continuously invest in having more unique data partnerships with key players in the market. This will strengthen our prediction power. Furthermore we really hope the general distrust in “big data” of the FMCG/ food industry will decrease. We see how data and data driven strategies has changed industries like film, music, and automobiles. We feel the time is now to embrace “data science” in the food & drink industry. That’s why we recently joined Esomar, the market research organisation. As an outsider in market research circles, we are excited at the prospective of challenging players in this sector to always be thinking of the next application of data.
Sign up for the free insideBIGDATA newsletter.