Before-and-after: two rulebooks that cannot be merged
Updated 2026-08-15 · Risk-term list version 2026-08-15c
These cannot be merged into one rule. TikTok’s misleading-content policy names "product effect comparisons, such as before-and-after results, which may cause viewers to have a false or distorted impression". Meta’s health and wellness standard permits general cosmetic before-and-after transformation for 18+ targeting, while barring "statements of inferiority about physical appearance" and a "close up on specific body area by pinching fat".
So the right answer is different cuts per platform: a single-variable demonstration on TikTok, and a comparison on Meta with targeting and framing handled.
Below: the two policies side by side, what "distorted impression" actually turns on, three formats that are safe on both, and a pre-shoot checklist.
1. The two policies side by side
Put the wording in parallel, because the difference lives in the wording. Collapsing both into one internal rule is the common mistake, and the cost is asymmetric: "never" throws away a working format on Meta, "always" gets you rejected repeatedly on TikTok.
| Platform | Key policy wording | Position on before-and-after |
|---|---|---|
| TikTok (misleading and false content, page states April 2026) | Bars "product effect comparisons, such as before-and-after results, which may cause viewers to have a false or distorted impression about a product's outcome/results", and separately bars ads that "promise or exaggerate results concerning a product's effect" | Names the format. The test is whether the impression is distorted, not whether there are two images |
| Meta (health and wellness) | General cosmetic "before and after transformation" is permitted with 18+ targeting; ads must not contain "statements of inferiority about physical appearance" or a "close up on specific body area by pinching fat" | Permitted, conditionally. What trips it is usually the line and the close-up, not the comparison |
| Meta (privacy violations and personal attributes) | Ads "must not contain content that asserts or implies personal attributes... physical or mental health (including medical conditions)" | A second-person body question is more dangerous than the comparison itself |
2. What "distorted impression" turns on: count your variables
TikTok gives no operational definition, but you can work backwards from the phrase. The real question is: how much of the difference the viewer sees was caused by the product?
A typical before-and-after changes several things at once: the outer garment, the pose, the lighting, the camera angle, an implied time span, whether the person is holding their breath, and retouching. The more variables move, the less defensible "the product did this" becomes.
| Variables changed | What the viewer reads | Risk |
|---|---|---|
| One (only the layer underneath; same outfit, camera, light, pose) | This garment smooths the line under clothes | Low. This is a demonstration, not an outcome claim |
| Two or three (different outer garment, or a different pose) | Unclear whether it was the garment | Medium. Starts to look like angle-picking |
| Four or more, or an implied time span ("two weeks later") | Wearing this changes your body | High. This is the outcome claim TikTok names |
| With a quantified result ("two sizes smaller", "lose five pounds") | A quantified body-change promise | Highest — and those phrases are themselves on the flagged list |
3. Three formats that are safe on both
Rather than litigating whether you can run a comparison, switch to a format that is easier to shoot, more credible, and clean on both platforms. All three below are the core shots we recommend on the angle pages:
- The single-variable demonstration. Same fitted outer garment, same camera, same light, only the layer underneath changes. Its persuasive force comes from exactly one thing having moved — the opposite of a distorted impression. This is the most effective shapewear shot there is, not merely the safest.
- Timestamped evidence. "After a full day of sitting" is not a body comparison; it is a comparison of the garment — did the waistband roll between morning and evening. Rolling is the top return driver in this category, so buyers genuinely care, and it is entirely about the product.
- Third-party framing. Do not give a result, give someone else’s words: "the most repeated line in the reviews is that you really cannot see it." The claim moves from something you promise to something buyers reported, and it carries social proof for free.
4. If you keep the comparison on Meta, handle three things
- Target 18+. Weight-loss and cosmetic ads require it, and comparison formats especially. This step is in the ad manager, not the edit.
- Delete the appearance-inferiority line. "Still struggling with your belly?" is the textbook violation. Replace it with a first-person account: "I sat in this for a full day."
- Delete the pinch-the-fat close-up. The standard names that frame. While you are there, delete scales, tape measures and body measurements too — they are not in that clause, but they push the asset into health review, which is a bad trade.
5. Chinese domestic platforms word this differently
Everything here is about the US market. Xiaohongshu, Douyin and Taobao each have their own rules on before-and-after and body-related content, layered on top of Chinese advertising law provisions on absolutes and false claims.
Do not carry the US conclusion home, or the other way round. We verified US platform policy pages only (linked at the bottom of this page), so this article makes no claim about the Chinese side — we do not write what we have not checked.
6. The pre-shoot checklist
- Besides the layer underneath, what else changed? (Outer garment, pose, light, camera, time.) More than one — rethink it.
- Any scales, tape measures, body measurements, or pinching fat in frame? Cut them.
- Is the line describing the garment or the body? Rewrite the body ones (method: the claims mapping).
- Any implied time span ("after two weeks", "keep wearing it")? That is an outcome claim. Cut it.
- Second-person question or first-person account? Convert questions into accounts.
- Is the TikTok cut a single-variable demonstration?
- Is the Meta cut targeted 18+?
- Have these lines been through the checker?
Related
FAQ
So do I shoot it or not?
Shoot the single-variable demonstration: same outer garment, same camera, one changed layer. It is clean on both platforms and it is also easier to shoot and more convincing — viewers believe "only one thing moved". If you want a traditional comparison on Meta, handle targeting and framing as in section 4 first.
Does a "results vary" caption save it?
Not as an exemption. Disclaimers are necessary — this category already requires statements like "this is a garment; it provides no weight-loss, medical or corrective effect" — but they do not change what the asset promises. Review reads the whole asset, not the small print.
What if the creator shoots one anyway?
Put "no before-and-after" in the do-not-show section of the brief and repeat it when you confirm the sample — the platform reviews the video that went live, regardless of who filmed it. Then gate the delivery with the acceptance scorecard, whose section A is a hard pass/fail.
Everyone else is running them. Why can I not?
You are seeing the ones that have not been actioned yet. Rejections, throttling and takedowns are not public, so "everyone is doing it" carries survivorship bias. Enforcement intensity also moves; fine this quarter is not fine next.
What about Amazon?
This article covers TikTok and Meta only, because those are the policy pages we verified line by line in this pass. Amazon’s ad specs are on the Amazon page — its aspect-ratio rule alone differs from everyone else — and we do not write about its claims stance without checking it.
Sources
- TikTok 广告政策 · 误导与虚假内容（页面自标 Last Updated: April 2026） （checked 2026-08-15）
- Meta 广告标准 · 健康与保健 （checked 2026-08-15）
- Meta 广告标准 · 隐私侵犯与个人属性 （checked 2026-08-15）