We want to make it easy for you to understand the environmental impact of your food. It’s important to us that the information we provide is transparent and reliable, as well as accessible and easy to understand. In addition to the total carbon footprint, we give each food a set of scores on a scale of 1-100 based on their emissions in relation to their nutritional content.

The carbon score compares the emissions of the product to the emissions from other products on the market—the higher the score, the lower the emissions per kilogram. (Yes, we use grams and kilograms for our calculations, but the results are the same if you convert to pounds).

Beside the carbon score, we have three efficiency scores: the protein score, the calcium score and the calories score. The protein score shows how efficient the product is as a source of protein—a high score means it has a lot of protein and relatively low emissions. The calcium score and calories score work the same way—they show how efficient the product is as a source of calcium or calories.

The problem with comparing the carbon footprint of different foods is that not all food is equivalent! We need a variety of foods to be healthy, and we use foods in different ways. For example, olive oil has a relatively high carbon footprint per kilogram compared to lettuce, but when’s the last time you drank a whole kilogram of olive oil? Or say you’re trying to build muscle and you want to find food that has a lot of protein—you want to understand the relative efficiency of food in terms of protein. The different scores let you compare the carbon footprint of foods in different ways, so you can build a more sustainable diet while still achieving health and fitness goals!

The carbon score is calculated along a nonlinear scale where the worst product (highest emissions) gets a 0 and the best product (lowest emissions) scores 100. We used an exponential scale because some products, like beef, are exponentially worse than others. This is the main score you see when you scan a product or add a recipe.

The efficiency scores (protein, calcium, calories) are calculated in relation to the best product (lowest emissions per nutritional unit) in our database. We use a logarithmic scale, so that every 10 points represents 2x the emissions. So a product with a protein score of 90 has twice the emissions per gram of protein of the “best” source of protein.

All of our nutritional data is from the USDA. We get carbon footprint data from multiple sources, including peer reviewed journal articles and reports from reputable international organizations. Our primary source of carbon footprint data is a comprehensive analysis done by Poore and Nemecek (2018). We also referred to the World Resources Institute Cool Foods Calculator, based on the same study, which expands the analysis to a number of common foods used in restaurants. Luckily, these sources alone cover most common foods in the western world! However, for foods that were not included, we utilized other published papers and associated supplementary data, ensuring that we were assessing similar factors. We also incorporated data on opportunity costs from Searchinger et al. 2018.

What about fish?

Estimating emissions for fish is HARD! See our blog post here. So much depends on where the fish is from, whether it was farmed or wild caught, and what methods were used. In the end, we relied on a number of sources, particularly research by Gephart, Henriksson, Parker, Shepon et al. that draws on data from over 2500 farms and fisheries around the world.

As with other foods, our data for fish does not include emissions relating to packaging, transport, or food waste. These stages typically represent a relatively low percentage of overall emissions, and vary hugely based on the specific product, location of purchase, and behavior of the final consumer (e.g. you and me!).

In cases where we know whether a seafood product uses wild caught or farmed fish, we can calculate emissions and scores more precisely. If we don’t know—which is a lot of the time—we assume the worst. That is, until we can partner with specific producers to verify the data!

Not all greenhouse gasses are created equal! Some contribute to climate change more than others. CO2e stands for “CO2 equivalent” or "Carbon Dioxide Equivalent" which converts the global warming potential of all greenhouse gasses to the equivalent of CO2.

To make it easy to read, we convert this data into “real life” examples, for instance, the number of miles driven and the number of phones charged. To do this, we use data from the EPA Greenhouse Gas Equivalencies Calculator.

Most carbon footprint assessments consider land use change, such as the emissions from cutting down trees to grow crops to feed cows. But what about feed that is grown on existing agricultural land? And how far back does the change have to be for it to be counted? What about land that was converted to agriculture 100 years ago?

We think (and many researchers agree) that a better measure is the difference between the net emissions of the current land use and the best use, such as a forest or the natural vegetation of the area. So if we weren’t using the land for agriculture, maybe we could plant trees and create a carbon sink. Using the land for agriculture instead of forest represents a lost opportunity, so the associated emissions are called opportunity costs.

Why is it better to consider opportunity costs instead of land use change? For one, it’s more accurate at showing the real impact of production. It makes it impossible for producers to hide behind labels like “deforestation free”. And it highlights the enormous opportunity for reducing emissions from agriculture.