Published 13th August 2025
ZOE’s new photologging app: Does it work?
Share this article
In this article, we’ll explain how photologging is solving two major problems people face when trying to eat healthily. Firstly, what product is best?
If you walk into any store, you can be forgiven for being overwhelmed by the choice of products.
We all want to eat well to support our health, but with so many options, it can be difficult to know what product is best.
Secondly, many of us want to track what we are eating as we aim for a balanced, diverse diet, but using food logging apps can be a pain — they require manual entry, which can be a pain.
Nutrition researchers face a similar struggle. If they want to understand how food impacts health, they often have to rely on laborious food diaries, which are time-consuming for participants to fill out and for scientists to analyze.
Food diaries are also prone to errors; none of us has a perfect memory, and some of us might be tempted to “accidentally” forget to add that second bag of chips.
Thankfully, nutrition science never stands still. Recently, photologging has entered the chat.
We don't buy the hype — and neither should you
Our new app reveals what the food labels won't, using data from the world's largest nutrition study run by ZOE. (Only available in the US)
What is photologging?
Thanks to swift advancements in artificial intelligence (AI) technology, photologging is growing in popularity.
Simply snapping a picture of product packaging, a food item, or a meal can provide nutritional information in seconds.
However, anyone who’s used an AI tool knows that they can sometimes be hit and miss:
Chatbots can hallucinate and give poor advice. And like all new tech, some versions are better than others.
As a science-first company that has recently released a free photologging app in the United States, we needed to investigate how accurate our technology is.
So, we ran a study and we recently presented the results at the American Society for Nutrition conference in Orlando, FL.
Investigating photologging
We analyzed 10,000 meals from 2,124 participants in our PREDICT 1 and PREDICT 2 cohorts, all of whom were based in the U.S. or the U.K.
Using this data, we found that ZOE’s photologging technology works well: It’s reliable and much faster than traditional methods. Photologging is the future!
For each meal, participants took a photo and provided a weighed food log — this involves weighing all food items. It’s a very time-consuming process and is considered the gold standard for logging food intake.
We created a score for each AI photologging response, which ranged from 0 (terrible) to 1 (perfect).
We found that the ZOE’s AI performed very well, particularly for:
tea and coffee: 0.88
vegetables: 0.87
baked beans 0.84
However, photologging performed less well for other groups, such as:
plant milk: 0.13
soup: 0.07
hot chocolate: 0.07
The reason for the low scores for plant milk and soup is that they’re difficult to discern even for human eyes. One plant milk looks very much like another.
Also, the AI is biased towards more common food items, which is why hot chocolate — a less common drink — could be generally mistaken for tea or coffee.
Because AI can be trained, personalization will help reduce these errors. For instance, over time, the algorithm will learn that an individual often has hot chocolate in the evening.
Photologging also estimates the macronutrients in meals. Overall, it did a great job. For instance, it only overestimated:
calories by 7%
carbs by 3%
protein by 1%
Quantity is difficult to assess without weighing things, but knowing there are minor inaccuracies like this means that we can adjust the model moving forward and keep improving it.
It’s true that weighed food logs would perform better than photologging, because every item is weighed individually; not something you could do for every meal.
But if our photologging app were compared with standard food diaries based on memory, it would likely be better. And much less hassle.
Our new app gives you the power to see beyond the marketing
Make smarter, science-backed food choices in seconds. Scan. Score. Reveal the truth. (Available only in the US)
What else does ZOE’s new free app do?
As we’ve seen, ZOE’s new free app identifies the components of a meal and the macronutrient composition: fats, protein, and carbohydrates.
You can also use it to photograph packaged items or scan their barcodes to provide the same information.
Each meal or product gets a score out of 100 to let you know how it is likely to influence your health if consumed regularly in the long run.
Alongside these scores, the app provides information and short lessons about how to support your gut and overall health.
The app also introduces Ziggie, your friendly and motivating in-app nutrition coach who makes it incredibly easy to understand complex science, celebrate your wins, and build new healthy habits that stick.
It’s like having a professor in your pocket.
By completing a quick quiz, the app will personalize your nutrition advice to help you reach the goals that are important to you.
Finally, our app includes the perfect feature for navigating your local grocery store: ZOE’s Processed Food Risk Scale.
ZOE Processed Food Risk Scale
Many studies show that consuming certain types of ultra-processed foods (UPFs) is associated with poorer health.
However, the latest evidence suggests that some UPFs have a neutral effect on health, and some can even be beneficial.
This makes choosing products even more challenging. That’s why we designed the ZOE Processed Food Risk Scale.
It takes into account important factors that influence whether a UPF affects health:
Energy density: How many calories per gram of food.
Energy intake rate: How quickly you consume the calories.
Hyperpalatability: This can lead to excess consumption.
Non-culinary additives: These include artificial colors, sweeteners, and preservatives. We accounted for whether the European Food Safety Authority (EFSA), the Food and Drug Administration (FDA), or other official bodies deem them neutral or harmful.
Using this information, the ZOE Processed Food Risk Scale assesses the likely health impact of processed foods. It gives the level of risk as:
Unprocessed: This food is in its original form or close to it. Processing does not impact how healthy it is.
No Risk: Minimal processing that poses no risk to health.
Low Risk: The processing of this food has no or very low impact on health.
Medium Risk: The processing of this food may make it less healthy and could have a moderate impact on your health.
Highest Risk: The processing of this food makes it less healthy and could have a high impact on your health.
Summary
ZOE’s scientists recently tested our new free app’s photologging capability and found it to be impressively accurate.
Our app can help guide you as you navigate our confusing modern food environment, teach you about nutrition, and personalize your nutrition journey.
Without needing to type in ingredients or make guesses based on product labels, you can get all the information you need to stay healthy by taking one photo.
If you live in the U.S., download the app for free here.
