AppLovin Ads for Beauty Brands: What Works in 2026

AppLovin ads work for beauty and skincare brands when three things are true: you can feed the platform a steady stream of creator-style video, each hero product has room to learn on its own, and every claim in the ad would survive both AppLovin's health and wellness policy and the FDA's line between a cosmetic and a drug. The public evidence is stronger for beauty than for most categories, with three published AppLovin beauty cases, but one independent geo test also shows where it can fall short on new customers.

This guide covers what that evidence says, how to structure campaigns, what to put in the video and the end card, which claims to keep out, and how to measure the result honestly. If you are new to the platform, start with our explainer on AppLovin ads.

What the public evidence shows

Beauty is one of the few ecommerce categories where you can read several AppLovin results side by side. The table summarizes what each source actually reports. None of these are Apt clients, and the first three are published by AppLovin itself, so read them as best cases, not averages.

Two patterns stand out. First, the brands that scaled fastest treated creative as a volume business, adding new ads every week or every day. Second, the one independent test is a useful counterweight: AppLovin drove incremental orders for Jones Road Beauty, but many of them came from returning customers. Haus noted at the time that brands could not exclude existing customers, which made new-customer acquisition harder to steer. If your goal is new buyers, build your test around that number from day one.

Why the format suits beauty products

An AppLovin ad usually plays inside a mobile game, full screen, as a portrait video followed by an interactive page and often a catalog screen. That sequence fits beauty for three reasons.

1. Demonstration sells the product. Texture, application and a visible before-and-after routine are easy to show in 15 to 60 seconds of vertical video, and rewarded placements can hold attention far longer than a social feed.
2. Price points suit impulse purchases. MAËLYS reported that 94% of AppLovin purchases happened within an hour of the click. That fits the Day 0 goal most beauty brands start with. If your products cost more or need longer consideration, read our guide to AppLovin Day 7 optimization before choosing a window.
3. Catalogs are deep. Kitsch sells hundreds of products, and the catalog screen lets the ad show related items after the hero video, which helps bundles and second purchases.

Campaign structure for beauty brands

Kitsch describes the clearest structure: a dedicated campaign per product, so each product builds its own learnings and can be scaled on its own as demand grows. Most of those campaigns were Discovery campaigns, which aim at people who have not visited your site before. Our guide to AppLovin Discovery campaigns explains the data each campaign type needs.

Besque ran a cost per purchase campaign alongside a ROAS campaign. The first aims to convert the most shoppers at a target cost, the second aims at higher-value orders. Running both is a sensible way to learn which goal fits your average order value before you commit budget to one.

Creative that works for beauty on AppLovin

AppLovin's video best practices page reviews top-spending ecommerce videos. Most of the patterns map directly onto beauty.

• Captions on everything. 95% of the top-spending videos had prominent captions, because most players watch with sound off. Caption the benefit, not the ingredient list.
• Voiceover and real people. 85% used voiceover, and the vast majority were UGC style. Kitsch's videos show real people using the product, with hooks such as cutting a 20-minute routine to 30 seconds or calling out frizzy, curly or wavy hair.
• Offers. About 40% of top videos carried a promotion. Besque tested lines such as "Buy 2 bottles, get 1 free" and "Try risk-free with free shipping".
• Audience framing. About 30% named the group the product suits. For beauty, name a hair type, a routine or a lifestyle, not a skin condition (see the policy section below).
• Length. AppLovin recommends testing all lengths with a focus on videos over 30 seconds, and stitching several creator clips together if you lack long-form footage.

Then make the volume. In AppLovin experiments reported by Common Thread Collective in June 2026, campaigns using all available creatives beat hand-picked sets of 5 to 10 top performers by 7 to 23 percent in every test, across meal delivery, supplements and beauty. Those all-creative campaigns averaged 182 ads. AppLovin also advises against pausing creatives manually, because the optimizer already shifts spend away from weak ones. For the file formats, see our AppLovin creative specs.

End cards and landing pages

Since June 2026 you can upload .png, .jpg or .gif end cards directly in Ads Manager, and a call to action is added automatically. That makes it cheap to follow Besque's rhythm of 2 to 3 new end cards a week, testing the offer, the bundle and the shade range. Besque also built landing pages for AppLovin traffic and tested exit-intent popups, FAQs and review videos. A shopper arriving from a game has seen 30 seconds of video and nothing else, so the page should repeat the video's promise and answer the obvious questions about skin type, shade and returns.

Policy: what beauty ads can and cannot say

AppLovin's restricted content policy for health, wellness and pharmaceutical products, updated February 27, 2026, lists cosmetics, shampoos and body cleansers among products not subject to FDA approval. Ads in this group need express approval from the AppLovin team, and the rules are close to those for supplements, which we cover in detail in our post on AppLovin ads for supplement brands. For beauty, the rules that matter most:

1. No medical claims or claims that could be read as medical, and no "prevent", "cure" or "treat".
2. Ads and landing pages carry an appropriate FDA or comparable regulatory disclaimer.
3. No before-and-after images showing unlikely or extreme results, and no focus on "problem areas" that may evoke negative self-perception.
4. No second-person language that implies you know the viewer's health status or conditions. "Got acne?" is out; "for oily skin routines" is safer.
5. No ads that exploit insecurities or promote negative body image, and no exaggerated benefits.
6. Claims need support, and you cannot target anyone under 18.

Under that sits the FDA's own line. A product's intended use, shown by its advertising claims, decides whether it is a cosmetic or a drug. The FDA lists claims such as restoring hair growth, reducing cellulite, increasing or decreasing melanin production and regenerating cells as drug claims, and treats acne cleansers and sunscreen-claim moisturizers as drugs, or both. A hair care brand that says "regrows thinning hair" in an ad has changed what it is selling in the eyes of the regulator. This is not legal advice; have someone qualified review your claims before launch.

How to measure beauty results honestly

Beauty brands usually sell through Amazon, Sephora, Ulta or Target as well as their own site. Kitsch, for example, sells through Target and Ulta Beauty. That means platform ROAS will miss some of the effect and click-based tools will over-credit some of it. Three habits help.

1. Separate new from returning orders. The Jones Road Beauty test is the reason. Total orders rose, but new-order cost missed target badly. Report new-customer CPA next to blended ROAS every week.
2. Run a geo holdout once spend is meaningful. Haus used holdouts of 20% to 50% of the US for two to three weeks, plus a post-treatment window. Our guide to AppLovin attribution and incrementality covers the setup.
3. Watch retail halo. If you sell in stores or on Amazon, track those channels in the test regions too, or you will undercount the result.

A 7-step launch plan for beauty brands

1. Confirm your claims, disclaimer and landing pages meet AppLovin's health and wellness policy, and request approval from the AppLovin team.
2. Install the pixel and server events, and check purchases deduplicate before spending.
3. Pick two or three hero products or bundles, each with its own campaign and a Day 0 ROAS or cost per purchase goal.
4. Launch with at least 15 to 20 captioned UGC videos per product, mixing lengths and creators, plus a clean catalog.
5. Add 3 to 5 new videos and 2 to 3 end cards each week. Do not pause losers by hand.
6. After two to three weeks of stable spend, report new-customer CPA and blended ROAS side by side.
7. Run a geo holdout before you scale past your test budget, then scale the products that beat target.

If you would rather have a team run this for you, see how our AppLovin agency works, or read more about AppLovin ads for beauty brands on our service page.

FAQ

Do AppLovin ads work for beauty brands?

Published AppLovin cases from MAËLYS, Kitsch and Besque report fast scale and ROAS at or above target. These are platform-published results. An independent Haus geo test on Jones Road Beauty found real lift in total orders but poor new-customer efficiency, so test with your own new-customer target.

Do beauty and skincare ads need approval on AppLovin?

AppLovin's health and wellness policy, updated February 27, 2026, lists cosmetics, shampoos and body cleansers among the products it covers and requires express approval from the AppLovin team for that content. Ask your account team before launch.

Can I use before-and-after photos in AppLovin beauty ads?

Only with care. The policy bans before-and-after images that show unlikely or extreme results or focus on problem areas in a way that may evoke negative self-perception. Realistic routine results with a disclaimer are the safer choice.

Which optimization goal should a beauty brand start with?

Most beauty brands start with Day 0 ROAS or cost per purchase. MAËLYS and Kitsch both optimized for Day 0, and MAËLYS reported 94% of purchases within an hour of the click. Use Day 7 for higher-priced kits or slower purchases.

How many creatives does a beauty brand need on AppLovin?

More than most expect. In AppLovin experiments reported by Common Thread Collective, campaigns with all available creatives, 182 on average, beat curated sets of 5 to 10 by 7 to 23 percent. Besque added 3 to 5 videos and 2 to 3 end cards a week.