Do certain farming practices yield more nutritious bounties? Does soil health contribute to nutrient density? Can eating more carrots really help me see better?
Consumers’ evolving interest in health and diet has sparked many questions of this nature, and the ongoing coronavirus pandemic is likely to throw them into even further into the spotlight as we learn more about the link between what we eat and how we fight viruses like Covid-19.
But measuring the nutrient density of foods is still a rare and expensive art, according to Greg Shewmaker, co-founder of food data company TeakOrigin. A few startups have created ways of evaluating nutrient density in food but the technology was due for an update, he adds.
“The only way to accurately understand a food’s quality is to have it analyzed in a properly equipped laboratory,” he recently told AFN. “The problems with that include prohibitive cost at roughly $900 per sample, it takes weeks to receive the results, it destroys the food in the process and creates waste, and the results aren’t aggregated or shared to help create larger insights into food quality.”
The Boston-based startup researches nutritional components that are key markers of quality in fresh foods. It uses this information to establish universal standards — a common language — for each of these foods, which it calls Critical Composition Profiles.
“Using scientific and data-driven insights, TeakOrigin establishes the true quality, value, and nutritional state of top-selling fruits and vegetables compared to known standards and expectations,” Shewmaker explains.
The precursor to TeakOrigin involved Target’s Food+Future initiative and a multi-year, highly-collaborative exploration into the future of food with dozens of global companies and institutions. This collaboration provided the startup with unparalleled insights and exposure to so many innovative companies. To date, it has also worked with Walmart, Amazon, Costco, Tesco, Coles, PepsiCo, Dole, Driscoll’s, and Ferrero.
Different food languages and quality controls have proved major problems for supply chains during the Covid-19 pandemic as farmer suppliers to the foodservice industry have been unable to transition their produce to the much in-need grocery industry that has different quality controls and a length on-boarding process before bringing on new suppliers. TeakOrigin is looking at how it can scale the use of its “language” across the sector to help.
Read on to learn more about the creation of TeakOrigin, how the tech works, and how the company is responding to Covid-19.
When did you launch and what growth stage are you currently at as a company?
We began this journey more than six years ago while working at Tesco in Hong Kong. When Tesco hit a rough patch, we moved our R&D efforts back to the US, partnering with Target, MIT, Intel, and others to develop a solution. These early efforts all focused on creating a hardware solution. However, we learned that the hardware was useless without accurate referential datasets, a common language of food quality that could be spoken in any situation.
In 2017, we spun TeakOrigin out as its own company to focus on building this common language for fresh foods. We are now conducting consumer pilots in Los Angeles and Boston and commercial pilots with two of the largest food retailers in the world.
How does your technology work?
Using a combination of analytical chemistry, optical spectroscopy, and machine learning, we build high-fidelity prediction models and digital datasets for each of these components contained in a specific food. We then transfer these models to low-cost, lower-fidelity field spectrometers. Our models enable these lower-fidelity field devices to operate like much-higher-fidelity lab instruments. The technology application works on a wide range of fresh foods, including produce, meat, seafood/fish, grains, herbs, oils, dairy, wine, etc. We have started by focusing on produce because it is where we had conducted the most R&D and early proof of concepts.
Right now, we are piloting two products—a free version for consumers and a paid version for food retailers. Every week TeakOrigin employees assess thousands of top fruits and vegetables purchased at a selection of ten retailers in Los Angeles and ten more in Boston. We publish the results each Friday in the TeakOrigin Guide, a free resource designed to unearth the unique inner qualities of food.
We launched the TeakOrigin Guide for two primary reasons. First, to help raise awareness to the fact that we’re not always getting what we’re paying for when it comes to produce quality (sometimes less than we expect, sometimes more than we expect). Second, to leverage this awareness to drive improvement back through the supply chain.
In addition, we are conducting a commercial pilot with a large retailer looking to predict the quality of top produce items prior to their arrival into its warehouses and stores and also to compare these foods to internal standards and to their competitors’ offerings. In this case, TeakOrigin is providing data services (tools and onsite support) and deep analytics into massive amounts of data.
The second retail pilot is much more consumer-focused. In this case, customers perceive the retailer’s produce to be of extremely low quality. Our data says something quite different. The retailer will work to bridge this gap between what is believed and what is real by sharing actual results with its customers and a very public campaign about its work with TeakOrigin to ensure it’s selling high-quality fruits and vegetables. TeakOrigin is providing data services (tools, onsite support, audits) and deep analytics into massive amounts of data.
What’s been the most surprising aspect of your startup journey so far?
When we first started this journey, we were absolutely convinced that the proverbial ‘man’ was sticking it to us all. We just knew that big food companies were cutting corners and selling us food that they knew wasn’t up to standards, or only fit all the marketing or branding.
Of course, there are some bad actors, but we’ve come to learn through the years that these are the exceptions. The vast majority of producers/resellers have no clue about the quality of food moving through their systems. They can’t. There’s no easy way for them to understand.
We’ve been with some of the most experienced executives and professionals, and the most senior officials from the world’s largest governments and have said, probably a thousand times, things like, “We know less about our food than ever before,” or “We have no idea about what’s actually happening inside the food we grow/sell/regulate/eat.”
Each time we wait for someone to tell us how crazy and off-base we are, that the answers we seek are already out there. But that has never happened, not once. Instead, these people usually just nod their heads in agreement. That’s been shocking to us.
How has Covid-19 impacted your business to-date?
To date, our business has been impacted in three different ways. First, early in the shutdown process, we made the decision to stop our daily collection of food quality data in Los Angeles and Boston grocery stores for the TeakOrigin Guide. This was done primarily to protect our team but also to make sure we weren’t taking food off the shelves that needed to go to families stocking up and preparing for quarantine. This also delayed our formal roll-out of the TeakOrigin Guide to those communities.
Next, we were preparing to launch a couple of multi-city pilots with a large US retailer. The goal was to help the retailer improve its customers’ perception of the retailer’s fresh offering. Obviously, COVID-19 shifted everyone’s focus to availability, so we jointly made the decision to delay these pilots until things begin to look somewhat normal again.
Lastly, as has been reported in the news, testing labs around the world have gone dark (at least those not working on Covid-19) and food inspectors have stopped field assessments on fresh foods moving through the global supply chain. We didn’t want this to stop our work building out and testing food models so we created a decentralized “lab@home” for TeakOrigin team members to continue performing lab-grade analysis of nutrition while sheltering in place at home.
Both scientists and non-scientists at our company took instruments home with them and have been analyzing the quality of fresh produce they’re buying and feeding their families. We’ve been able to share some of these results with producers and retailers and plan to share more information directly with other consumers in the coming weeks.
Does your technology offer a solution to help ease the impact of Covid-19?
Covid-19 is definitely setting off panic within the food system. Here’s a few intense examples of what we’re experiencing: On Monday an executive from one of the world’s largest retailers called and asked if we could help them diversify their food supply chain to meet existing demand and to take advantage of new producers proactively reaching out with supply. The person explained that their existing supply chain was extremely efficient but as the crisis hit, it was struggling to shift in other directions to keep food on the shelves.
On Tuesday another large retailer called and echoed the same problem. This time the executive explained how they’d spent years building a highly strategic supply chain, with far fewer suppliers. The problem is, that now as they work to keep fresh food flowing, they are struggling to qualify and add new suppliers located across the globe.
In between those two calls, I also spoke with two producers (one small, one large) that both had food rotting in the field because their traditional customers/channels had completely shut down.
So, on the one hand, we have businesses with too much food and then others with not enough. That’s our food system! I spent most of my professional life in international sourcing/merchandising. In non-food businesses, we always had robust databases (often shared with ‘competitors’) of qualified suppliers and what their specialties were, when they were last socially audited, published product specifications, testing results, etc. Nothing like this exists for fresh foods. Sure there are plenty of lists of producers and marketplace platforms that connect buyers/sellers. But because food is so dynamic, changing from harvest to harvest, shipment to shipment, etc. it’s been impossible to create and maintain an industry-wide, shared resource like this.
In each of the conversations above and others, we talked about building a comprehensive shared tool using TeakOrigin as a ‘common language’ that enables retail buyers to clearly establish standards for what they’re looking for, producers to quickly/cheaply prove that their products meet or exceed these standards, and for dynamic cost-pricing to be possible based on actual quality versus these standards.
Putting something like this quickly into place could give retailers greater flexibility and peace of mind in keeping their shelves full while giving suppliers that don’t normally have a seat at certain tables, the ability to communicate with the market, not through networking skills, but by directly communicating the quality of their fresh foods. This could benefit consumers that want the best foods, at the best value, retailers trying to maintain supply with fewer resources, or in a virtual manner, and producers trying to survive.
Guest article: AI can transform precision agriculture, but what are the legal risks?