As a small enterprise, Tribe 9 Foods was in search of a system that would provide more insights into how customers viewed its products without spending much money acquiring that information.
The food manufacturing company specializes in pasta and gluten-free foods. Its brand includes RP's Pasta, Taste Republic, Yumbutter and Ona Foods. With fewer than 50 employees, Tribe 9 Foods wanted to be able to allocate both its money and resources into the right technology.
"Sometimes you can get away with one or the other, and spend a bunch of money if you don't have people, and that will get your efficiency for you -- or spend a little money, but put a ton of work into it to make up for that," said Andrew Allcock, senior marketing director at Tribe 9 Foods. "I can't do either."
Native AI's technology
In search of a tool that can provide information on what consumers are thinking while still meeting its cost and talent requirements, Tribe 9 Foods, based in Madison, Wis., came across Native AI.
Native AI is an AI startup that pulls data from public data sets such as reviews or comments posted on Amazon. It uses that data to create a digital twin or synthetic user that enterprise customers can use to learn more about what consumers think of their products. The startup is careful to stay away from ingesting personally identifiable information by tokenizing that information through a server to Amazon's S3 bucket, a storage service.
The vendor seeks to provide enterprises with the same knowledge that brand managers or insight managers gain when researching what consumers think about certain products with surveys, chatbot panels and research projects.
"As we looked at something like natural language processing and also the sheer amount of data consumers are leaving around products or services online ... we knew there was a right way to use that AI to accelerate that time to insight," said Frank Pica, co-founder and CEO of Native AI. "Using natural language processing -- AI -- instead of human beings really reduced that cost."
Andrew AllcockSenior marketing director, Tribe 9 Foods
In Tribe 9 Foods' case, Native AI scraped all the reviews about the food manufacturing company from a public database. It imported that information into its platform, where Tribe 9 Foods could see consumers' consolidated sentiments about different product textures, flavors, packaging and other attributes.
This allowed marketers to dive into the reviews and know what consumers were saying without having to do the manual work of looking through the reviews themselves.
"[It] allows us to instantly always have a pulse on something," Allcock said.
Digital twins and synthetic users
Most recently, Native AI's digital twin technology has enabled Tribe 9 Foods to be more proactive instead of reactive about what consumers might think of new products the enterprise is considering.
Using the product reviews, Native AI generates a digital twin of a consumer that could be interested in the product that a customer such as Tribe 9 Foods releases.
For example, if Tribe 9 Foods wanted to release a pasta product, it could ask the digital twin what consumers think of the product or what food ingredient might make a difference. A question the manufacturer has asked the system is what's the difference between fresh and dry pasta.
"For us, that's really valuable because consumer research is expensive [and] takes a fair amount of time," Allcock said. "To be able to do something here and get pretty close to what that would be, almost instantly, is a huge advantage for us that allows us to compete with bigger brands that have bigger budgets."
As with most recommendation systems such as the one Native AI offers, challenges usually arise about whether the system provides the correct answer.
For Tribe 9 Foods, trusting the technology is about faith and business judgment.
"The only way to truly validate every single one of those [responses] would then be to go and run that actual research with consumers, which I can't afford to do," Allcock said. "So I have to take it with faith."
However, if the digital twin comes up with an answer that seems out of character with the responses the enterprise is used to getting, he added that the company would do more research.
"It's one of those where you can't trust anything blindly," Allcock said. Therefore, in those situations, the company's team works on phrasing more targeted inquiries to the synthetic user that provide valuable and efficient answers, he said.
Moreover, the enterprise has had instances where the AI technology coded a review incorrectly based on different slang the reviewer might have used. Therefore, the team had to go back and change the code from either negative or positive.
Despite some challenges with the technology, cost and time constraints have played a role in Tribe 9 Foods' decision to not research alternative tools, Allcock said. Native AI's competitors include Forestreet and Crayon.
The food manufacturer pays a discounted price of $1,500 per month for the technology and sees it as a necessary risk to stay competitive.
"All of this is pretty new, and a lot of people may be hesitant to use some of this," Allcock added. "We do some of it out of necessity, but we also do it because taking a little more risk on some of that can give us a competitive advantage over someone that's not willing to take that risk and try stuff that's new."
Esther Ajao is a news writer covering artificial intelligence software and systems.