60% of food and drink ads on Facebook and Instagram make a nutrition or health claim. Only 2% lead with indulgence.
We analyzed 930 active Facebook and Instagram ads from 24 US food and drink brands. In our sample, most mention protein, sugar or clean ingredients, and more than half of those still show the food being eaten or in mouth-watering close-up.

The big idea: In our sample, food and drink brands tend to put what's in the product into the ad and let the pictures carry the pleasure. One reading is that food science now lets health and indulgence sit in the same product, so brands no longer feel they have to choose. Consumer research suggests that pairing is essential because a health message on its own can make food sound less tasty. If there's an open lane, our data points less to indulgence and more to minimal processing: only 11% of ads talk about real or minimally processed ingredients, while about half of Americans say they're trying to eat "real food".
Protein, zero sugar, and gut health fill grocery shelves, and in our sample, they fill the ads too. We watched 930 of them to see what food and drink brands put first, what they show, and where they may be leaving room for a competitor. What follows is a pattern in one snapshot of ads active at the time of this writing.
Scope: Meta only. All 930 ads come from the Meta Ad Library, so they ran on Facebook, Instagram or both. TikTok, YouTube, Pinterest, Snapchat, retail media and TV aren't included. Brands often build different creative for each channel, so read these patterns as how food and drink brands advertise on Meta. They may not hold on other platforms.
We watched 930 active ads from 24 food and drink brands
We collected up to 45 active US ads per brand for 24 food and drink brands from the Meta Ad Library on November 6, 2026. For each ad, we reviewed the image or the full video with sound, plus the copy, headline, and call to action.
515 video ads and 415 image ads. All of them run on Instagram, and 93% also on Facebook.
Five product groups: beverages (41% of ads), snacks and sweets (18%), grocery delivery (16%), pantry staples (12%) and meals or meal kits (9%).
Group | Brands |
|---|---|
Beverages | Poppi, Liquid Death, Athletic Brewing, Celsius, Ghia, Liquid I.V., Oatly, Cometeer, Trade Coffee |
Snacks and sweets | Chomps, Mid-Day Squares, Legendary Foods, Tony's Chocolonely |
Pantry | Graza, Fly By Jing, Bachan's, Rao's |
Meals and delivery | Huel, Ka'Chava, HelloFresh, Thrive Market, Misfits Market, Snake River Farms, ButcherBox |
Sample sizes: most brands contribute 40 to 45 ads. Four had fewer active ads on the day: Legendary Foods (18), Ghia (17), Rao's (6) and Oatly (2). We keep them in the totals and flag them wherever we compare brands.
Your move: before you read on, try to guess how many of your own active ads mention protein, sugar or ingredients. Then count them.
In our sample, six in ten food and drink ads make a nutrition or health claim
60% of ads make at least one nutrition or health claim. 560 of 930 ads · 16 of 24 brands do it in most of their ads
Meta Ad Library, active US ads collected November 6, 2026. The percentage represents the share of ads.
We counted any claim about what the product does for the body or what goes into it: grams of protein, sugar content, clean or organic ingredients, gut health, energy, hydration, sourcing, calories and alcohol-free. The result barely changes when we leave out single brands at random: it stays between 58% and 63%.
Ingredient and sourcing language is the most common type, followed by functional benefits, sugar and protein.
Most claim ads stack them: 86% make two or more claims.
Ingredients and clean label: 27% (248 of 930 ads).
View data table
| Measure | Share | Ads | Sample |
|---|---|---|---|
| Ingredients and clean label | 27% | 248 | 930 |
| Functional benefits | 21% | 196 | 930 |
| Low or zero sugar | 21% | 194 | 930 |
| Protein | 21% | 191 | 930 |
| Fiber and gut health | 15% | 140 | 930 |
| Animal sourcing | 10% | 89 | 930 |
| Calories or carbs | 6% | 58 | 930 |
| Alcohol-free | 5% | 50 | 930 |
9AM analysis of 930 active US Meta ads from 24 food and drink brands, collected November 6, 2026. Percentages are rounded as published.
The claims cluster where you'd expect and in one place you might not:
Grocery delivery: 80% of ads. Thrive Market and Misfits Market sell the clean-label aisle itself.
Beverages: 67%. Prebiotics, electrolytes and zero sugar.
Snacks and sweets: 67%. Protein bars, meat sticks and a 20g-protein donut sit next to chocolate.
Meals: 42%. Pantry: 26%. Sauces and oils sell flavor instead.
Some claims go well past nutrition. 41 of Celsius's 45 ads say the drink accelerates metabolism or burns body fat (these are the advertiser's own claims).
The claims track what shoppers say they want.
Protein is the nutrient Americans most often say they're trying to get more of: 70% in 2025, up from 59% in 2022.
A high-protein diet has been the most common diet Americans report for three years running (IFIC, 2025). In Google Trends, US searches for "protein snacks" reached their five-year high in April 2026.
Your move: list the claims in your active ads and check how many competitors make the same ones. A claim every brand makes stops telling yours apart.
More than half of claim ads still show the food being enjoyed, mostly in video
54% of the ads that make a claim also show the product being eaten or in close-up. 303 of 560 claim ads · video: 71% (207 of 293) · images: 36% (96 of 267)
Meta Ad Library, active US ads collected November 6, 2026. "Shown being enjoyed" means someone eats, drinks, pours or cooks with the product, or the ad uses close-ups of texture, pouring or sizzling.
Video · with a claim: 71% (207 of 293 ads).
View data table
| Measure | Share | Ads | Sample |
|---|---|---|---|
| Video · with a claim | 71% | 207 | 293 |
| Video · no claim | 71% | 158 | 222 |
| Image · with a claim | 36% | 96 | 267 |
| Image · no claim | 43% | 63 | 148 |
9AM analysis of 930 active US Meta ads from 24 food and drink brands, collected November 6, 2026. Percentages are rounded as published.
A claim in the headline doesn't mean the ad gives up on appetite.
A Legendary Foods video promises 20g of protein and under 1g of sugar while maple syrup pours over a glazed donut. A Poppi video lists prebiotics and 5g of sugar over a macro shot of a frosted can.
The words make the case for the body, and the pictures make the case for the mouth.
The pattern holds across brands: 15 of the 21 brands that make claims show the product being enjoyed in at least half of their claim ads.
Format decides how much room there is for it.
Video showcases pleasure more vividly.
A static image usually has space for the pack and the claim, and little else.
Claims don't push the pleasure out, at least in video.
Video ads with and without a claim show the product being enjoyed equally often (71%).
In images, claim ads do it a little less: 36% against 43% for images with no claim. When space is tight, the claim tends to take it.
Industry trackers describe the same shift from the product side.
They call it "permissible indulgence": treats reformulated to deliver both enjoyment and nourishment (FoodNavigator, 2026).
Innova Market Insights found that 81% of global consumers snack to lift their mood, while 75% look for snacks with added health benefits (FoodNavigator, 2025).
The Legendary Foods donut, a glazed treat with 20g of protein, is that idea in a single ad.
See the Legendary Foods donut ad in the Meta Ad Library
In our sample, 23% of claim ads (128 of 560) lead with taste or indulgence and use the claim as support.
Your move: for every static ad with a nutrition claim, check whether anything in the image makes the product look good to eat. If nothing does, test a version with a texture shot or a hand reaching for it.
Taste leads more ads than health, and very few lead with indulgence
Only 2% of food and drink ads lead with indulgence. 21 of 930 ads · from 7 brands, led by Tony's Chocolonely with 6
Meta Ad Library, active US ads collected November 6, 2026. One main message per ad.
A claim somewhere in the ad is different from the message the ad leads with.
When we look only at the main message, taste comes first in 29% of ads (268 of 930) and health in 22% (200).
Novelty, convenience and price follow.
Pure indulgence, the treat for its own sake, is almost absent.
Its share stays between 2% and 2.4% whichever brand we leave out.
Taste: 29% (268 of 930 ads).
View data table
| Measure | Share | Ads | Sample |
|---|---|---|---|
| Taste | 29% | 268 | 930 |
| Health | 22% | 200 | 930 |
| Novelty | 16% | 151 | 930 |
| Convenience | 12% | 111 | 930 |
| Price or value | 11% | 102 | 930 |
| Social occasions | 4% | 38 | 930 |
| Sustainability | 4% | 36 | 930 |
| Indulgence | 2% | 21 | 930 |
9AM analysis of 930 active US Meta ads from 24 food and drink brands, collected November 6, 2026. Percentages are rounded as published.
The split depends on the category.
Pantry brands lead with taste in 45% of their ads and beverages in 38%.
Snacks and sweets lead with health (33%) more often than taste (21%), and with indulgence in only 5%.
Claim ads lean further toward health: 36% of the 560 lead with it.
Another 22% lead with taste and add the claim as support, like Celsius's watermelon can balanced on fresh fruit with "zero sugar" and "essential energy" beside it.
See the Celsius watermelon ad in the Meta Ad Library
Your move: if your category leads with health, test one ad that leads with how the product tastes and keeps the claim as the second line. In snacks, where health leads more ads than taste, that's the less crowded position.
Some brands claim in every ad, others never do
Liquid I.V., Huel and Legendary Foods make a claim in every ad. Cometeer, Trade Coffee and Tony's Chocolonely never do. 16 of 24 brands make a claim in most of their ads
Meta Ad Library, active US ads collected November 6, 2026. Oatly (2 ads) and Rao's (6) are too small to read on their own.
The sector average hides two camps.
Functional and protein brands claim in almost every ad: Chomps and Mid-Day Squares in 98%, ButcherBox and Celsius in 96%, Ka'Chava in 91%.
Brands that sell craft, origin or flavor rarely do.
Coffee and chocolate sit at zero, and Bachan's sauce at 10%.
Cometeer: 0% of 45 sampled ads make a claim.
View data table
| Brand | Share | Sample |
|---|---|---|
| Cometeer | 0% | 45 |
| Tony’s Chocolonely | 0% | 45 |
| Trade Coffee | 0% | 45 |
| Bachan’s | 10% | 40 |
| HelloFresh | 11% | 45 |
| Snake River Farms | 20% | 45 |
| Fly By Jing | 28% | 40 |
| Graza | 47% | 45 |
| Oatly | 50% | 2 |
| Rao’s | 50% | 6 |
| Liquid Death | 53% | 45 |
| Misfits Market | 62% | 45 |
| Athletic Brewing | 80% | 45 |
| Ghia | 82% | 17 |
| Poppi | 88% | 42 |
| Ka’Chava | 91% | 45 |
| ButcherBox | 96% | 45 |
| Celsius | 96% | 45 |
| Chomps | 98% | 45 |
| Mid-Day Squares | 98% | 45 |
| Thrive Market | 98% | 45 |
| Legendary Foods | 100% | 18 |
| Liquid I.V. | 100% | 45 |
| Huel | 100% | 45 |
9AM analysis of 930 active US Meta ads from 24 food and drink brands, collected November 6, 2026. Percentages are rounded as published. Dashed marker: all ads, 60%.
Equal weighting tells the same story. If every brand counts the same, the average is 61%, close to the 60% across all ads.
When we look at each brand's most common lead message, taste leads at 8 brands and health at 7.
Where the open lane may be. Most claims in our sample add something to the product: protein, fiber, electrolytes, or prebiotics. Language about real or minimally processed food is rarer. It appears in 11% of ads (99 of 930), and three in four of those pair it with a protein or sugar claim. Shoppers seem to be moving in that direction. In IFIC's 2026 Food & Health Survey, 52% of Americans recognized the term "ultra-processed", up from 32% two years earlier, and 49% said they had tried to "eat real food" in the past year (Dairy Processing, 2026). The January 2026 Dietary Guidelines for Americans also advise limiting highly processed foods. That suggests a counter-position built on minimal processing, food that's good for you because little was done to it, may stand out more than one built on indulgence.
Your move: map your direct competitors on the same line. If they all claim in most of their ads, a brand that leads with flavor has an open lane. If they rarely claim, a credible statement can stand out.
Consumer research suggests a health message can make food sound less tasty
Our data shows what brands do. It can't show what works, because the Ad Library has no spend or sales data.
Consumer research fills part of that rift, and it points to a real tension in the 60% figure.
60%+The unhealthy = tasty intuition
Over 60% of Americans in Raghunathan’s studies held this belief. Guests told a mango lassi was healthy rated it less tasty.
Raghunathan, Naylor & Hoyer · Journal of Marketing, 2006+29%Taste-focused vegetable names
Taste-focused names lifted vegetable selection by 29% versus health-focused names across 137,842 decisions in five dining halls. The effect held where the food tasted good.
Turnwald et al. · Psychological Science, 2019VariesHealth and taste are personal
European studies find that the link varies by category and person. Some shoppers expect healthy food to taste better.
Haasova & Florack · Food Quality and Preference, 2019More desireShowing the act of eating
People wanted food more and ate more when an ad showed a model engaged in eating rather than finishing.
Palcu, Haasova & Florack · Appetite, 2019Many shoppers assume healthy food tastes worse.
Raj Raghunathan, Rebecca Naylor and Wayne Hoyer called this the "unhealthy = tasty" intuition (Journal of Marketing, 2006).
Raghunathan has written that over 60% of Americans in their studies held the belief.
In one experiment, guests who were told a mango lassi was healthy rated it less tasty than guests told it was unhealthy.
Even people who said they didn't hold the belief showed the effect.
Leading with taste can sell the healthy option better than leading with health.
In a study across five US university dining halls and 137,842 diner decisions, Turnwald and colleagues (Psychological Science, 2019) found that taste-focused names like "Twisted Citrus Glazed Carrots" lifted vegetable selection by 29% compared with health-focused names.
Note: The effect only held where the food actually tasted good.
The belief isn't universal.
Studies with European consumers find that the link between health and taste varies by product category and by person, and some shoppers expect healthy food to taste better (Haasova and Florack, Food Quality and Preference, 2019).
In other words, a claim can help with one buyer and hurt with another.
Showing someone about to eat raises desire.
In two experiments, people wanted the food more and ate more of it when an ad showed a model engaged in eating, compared with a model finishing (Palcu, Haasova and Florack, Appetite, 2019).
Put together, the research suggests a nutrition claim on its own can lower how good the product sounds. Sensory proof, like the close-ups and bites in section 03, may offset that.
Remember: It's a hypothesis for your own tests, and our study doesn't prove it.
Seven more things 930 food and drink ads have in common
The claim story is one thread. These are the patterns a creative or media team can use right away.
Pattern | What we found |
|---|---|
Video shows the eating, images don't | 67% of video ads show someone eating, drinking, pouring or cooking with the product (343 of 515). Only 8% of image ads do (32 of 415). |
Snacks are the exception in video | Snack and sweet videos show the product being eaten in 42% of cases (36 of 85), well below the 67% average. |
Video looks like creators, images look like brands | 64% of video ads use a creator-style delivery (328 of 515). 86% of image ads are brand-led (358 of 415). |
The product shows up fast | In 82% of videos the product is on screen within 3 seconds (419 of 513). |
Most videos work on mute | 77% of videos still make sense without sound (394 of 515). The median video runs 31 seconds. |
Bold claims open the most ads | 31% of ads open on a bold claim, 14% on a product reveal and 10% on a taste close-up. In video, a person talking to camera opens 20%. |
Two in three ads have no offer | 66% carry no discount, bundle or free shipping (615 of 930). Discounts appear in 17%. |
One caution on creator-style: it describes how an ad looks and sounds. It doesn't tell us who made the ad or whether they were paid.
Your move: if your snack videos only show the pack and the macros, test a cut where someone takes the first bite in the opening seconds.
Six ads that show the range
Disclaimer: The claims below are the advertisers' own, subject to their conditions. Ads may stop running after our snapshot.
Legendary Foods, video: the claim and the craving together
Maple syrup pours over a glazed donut in close-up, and the copy promises 20g of protein and under 1g of sugar. The ad leads with taste and lets the macros reassure.
Poppi, video: function on a frosted can
A macro shot of a cold cranberry can, condensation and all, with "prebiotics" and "5g sugar" in the copy. Someone drinks it before the end.
Chomps, video: the dietitian's meal prep
A creator introduced as a dietitian walks through her fall meal prep and slips a meat stick into the routine. The ad leads with health: 10g of protein, 0g of sugar.
Celsius, image: fruit first, metabolism second
A chilled can sits on a stack of watermelon slices. The visual sells flavor, while the claims go further than most in the sample: essential energy, accelerated metabolism, body fat burning.
Tony's Chocolonely, video: indulgence without apology
A scoop of vanilla ice cream melts into a s'mores skillet of chocolate and marshmallow. No claim, no macros. It's one of only 21 ads in the sample that lead with indulgence.
Sauce drizzles over grilled chicken and skewers, and someone takes a bite. Bachan's leads with taste in 35 of its 40 ads and makes a claim in only 10%.
Four tests for food and drink teams, and what other categories can borrow
The data tells you where the feed is crowded. Your own tests tell you what works for your customers.
Lead with taste, keep the claim second. Take a claim-led ad and run a version that opens on flavor, texture or the first bite, with the same claim in the copy. Measure cost per acquisition and repeat purchase. Don't focus only on click-through rates because they don't tell you whether conversions are actually happening.
Give every static claim a reason to crave. Add a texture close-up, a pour or a hand reaching for the product to your image ads with nutrition claims. Compare it with the pack-and-claim version.
Show the bite early in snack videos. Snack videos show the product being eaten in 42% of cases. Test a cut where someone takes the first bite in the opening seconds, against your current opener.
If your product is simple, test leading with that. If your ingredient list is short, try an ad that leads with how little is in the product instead of what's been added. Keep the format and offer the same, and compare it with your current claim-led ad.
Before you launch, agree on the main outcome, the smallest improvement worth acting on and when you'll stop the test.
Beyond food and drink
The same tension between proof and pleasure runs through other categories. This part is our interpretation, and the study didn't measure these markets.
Food and drink
The proof they sellProtein, zero sugar, clean label
The pleasure to showThe bite, the pour, the sizzle
Supplements and wellness
The proof they sellMilligrams, percentages, outcomes
The pleasure to showThe morning scoop, the fizz in the glass
Mattresses and sleep
The proof they sellBetter sleep, less back pain
The pleasure to showSinking into the bed
Skincare
The proof they sellActive ingredients, percentages
The pleasure to showTexture and application on skin
Non-alcoholic drinks
The proof they sellNo alcohol, no hangover
The pleasure to showThe clink, the pour, the evening
Food and drink is based on this study of 930 ads. The other categories are illustrative.
Supplements and functional wellness sell outcomes in milligrams and percentages. Brands that also show the ritual, like the morning scoop or the fizz in the glass, give the claim a feeling to attach to.
Mattresses and sleep sell health (better sleep, less back pain) and comfort (sinking into the bed). Ads that show only specs and ratings risk the same trap as a protein bar sold on macros alone.
Skincare leans on active ingredients and percentages. Texture shots and application on skin do the work that close-ups do for food.
Non-alcoholic drinks sell the absence of alcohol. Ads that show the clink, the pour and the evening around the glass sell what the drink adds.
The common lesson: a claim answers "is it good for me?", and the creative still has to answer "will I enjoy it?".
Want help building and testing these ads? Talk to the 9AM team.
Sources and scope
What we collected. 930 active ads listed in the US Meta Ad Library for 24 food and drink brands, collected on November 6, 2026. We took up to 45 ads per brand from each brand's own page, in the order the Ad Library returned them: 515 video ads and 415 image ads. The library filter doesn't show who the ads were actually shown to. The Meta Ad Library covers Meta's own platforms only, so this study says nothing about the same brands' ads on TikTok, YouTube, Pinterest, Snapchat, retail media or TV.
How the ads were coded. An AI model reviewed every ad's image or full video, with sound, together with its copy, headline and call to action. It filled the same fixed set of fields for each ad: product, main message, opening, claims, offer, style, and whether the product is eaten or shown in close-up. Every label had to be backed by an observation, with a timestamp for video. We spot-checked a handful of ads by hand. There was no full human review of every label.
What counts as a claim. Any statement about nutrition, ingredients, sourcing, a functional benefit or the absence of alcohol, made anywhere in the ad.
What the numbers describe. Share of ads, not share of spend, impressions or sales. Each brand contributes up to 45 ads, so brands with fewer active ads weigh less. Where it matters, we also report brand-level counts and checks that leave out one brand at a time.
What the numbers don't show. We collected no spend, reach, conversion or sales data, so nothing here says which ads work. The 24 brands were chosen deliberately and don't represent the whole category. Ad claims are the advertisers' own and weren't verified.
FAQ
How many food and drink ads make health claims?
In a November 2026 sample of 930 active US Facebook and Instagram ads from 24 food and drink brands, 560 (60%) made at least one nutrition or health claim, such as protein, low sugar, organic or added functional benefits.
What do food and drink ads lead with?
Taste is the most common main message, in 29% of ads. Health leads in 22%, novelty in 16% and convenience in 12%. Indulgence leads in only 2%.
Do food ads show people eating?
In video, yes: 67% of video ads show someone eating, drinking, pouring or cooking with the product. In static images, only 8% do.
Do health claims make food ads less appetizing?
The study doesn't measure results. Consumer research suggests a health message can lower expected taste for some buyers, which makes sensory proof in the creative worth testing.
Research cited.
Raghunathan, R., Naylor, R. W., & Hoyer, W. D. (2006). The unhealthy = tasty intuition and its effects on taste inferences, enjoyment, and choice of food products. Journal of Marketing, 70(4), 170–184. Summary by the author: Supersizing Taste, ISBInsight.
Turnwald, B. P., et al. (2019). Increasing vegetable intake by emphasizing tasty and enjoyable attributes. Psychological Science, 30(11), 1603–1615.
Haasova, S., & Florack, A. (2019). Practicing the (un)healthy = tasty intuition. Food Quality and Preference, 75, 39–53.
Palcu, J., Haasova, S., & Florack, A. (2019). Advertising models in the act of eating. Appetite, 139, 59–66.
International Food Information Council. (2025). IFIC Spotlight Survey: Americans' Perceptions of Protein, with results from the 2025 IFIC Food & Health Survey.
International Food Information Council. (2026). 2026 IFIC Food & Health Survey, as reported in Dairy Processing, July 2026.
FoodNavigator. (2026). Permissible indulgence trend powers better-for-you boom. (2025). Top 10 snack trends for 2025, citing Innova Market Insights.
Google Trends, US web search interest for "protein snacks", November 2021 to November 2026.
Suggested citation: 9AM, Food and Drink Ad Study, November 2026. 930 active US Meta ads from 24 brands.