Dive Brief:
- Gastrograph AI closed a $4 million Series A round funding, according to AgFunder News. The company collects data about how people perceive flavors and then uses artificial intelligence to offer food companies predictions and suggestions on how a demographic might respond to products.
- Data for the system is collected by hundreds of professional tasters who class 24 variables about a food’s aroma, flavor and texture.
- However, even the CEO of Gastrograph AI does not believe that machines can replace humans in a taste test. “The resolution we can collect from each taster depends on the individual,” Jason Cohen, the company's CEO, told AgFunderNews. “We’re based in New York City, so we’re not going to have a representative set of tasters for the entire world, but we take data from the tasters we do have, and use that to predict what other demographics will perceive in a product.”
Dive Insight:
By taking tasting data from hundreds of human testers, Gastrograph is essentially becoming a storage bank of taste buds and underscoring the importance of human beings in the food industry.
As a New York-based startup, the tasters involved have skewed the company's databank toward a more American palate. After all, different cultures and demographics will perceive the same food in different ways because they have specialized food preferences. Interestingly, more than 75% of Gastrograph’s customers are outside the US and 50% are in Asia, AgFunderNews said. In fact, the latest Series A funding round was led primarily by Asian-based investors.
There is plenty of demand for artificial intelligence in the U.S. that is shaping and personalizing the modern food industry. In Los Angeles, Halla’s I/O platform operates much like Netflix – but is based on a database of taste and flavor attributes from local food hubs – to recommend individuals grocery stores, restaurants and food delivery platforms. Similarly, startups like Foodpairing, Plant Jammer, and Dishq also focus on making food-related recommendations based on a person's preferences.
On the manufacturing side, the landscape is different. Although coveted, artificial intelligence talent is so desirable that technology companies are spending more than $650 million annually to entice desirable candidates. With fierce competition and razor-thin margins, CPG companies don’t have the deep pockets of the tech giants who are developing AI, which means that the likelihood of in-house flavor-predictive artificial intelligence catching on is slim.
That leaves Big Food to rely on traditional focus groups and crowdsourcing for development. At the same time, large CPG companies are feeling the backlash of consumers who are shifting away from global conglomerates toward local, artisanal providers. In addition, consumers, especially millennials and other young shoppers, are looking for new flavors and variety in the foods they eat. Perhaps being able to mimic the small-scale adaptability of food startups through third-party artificial intelligence taste preferences predictions is the way to keep consumers engaged.
One interesting flavor preference that Gastrograph unearthed, according to Venture Beat, is U.S. consumers’ growing preference for sour food and drink (think IPA beer and grapefruit).
With new insight into real-time flavor priorities of their customers, CPG companies may put themselves on the track toward innovation while cutting costs and staying current. Doing so may help boost the sluggish growth that the industry has experienced. From 2013 to 2016, the industry grew less than 1.8% annually, on average.
Still, because it all comes back to human predilection, Gastrograph will need to assemble data based on different demographics or geographies. For the technology to work and be useful, the company generally can't just issue a blanket statement, assuming it will work across a broader segment of the popular. In all likelihood, Gastrograph and artificial intelligence will be one of many tools that food companies will use to help determine the success of a product with shoppers.