Sample creative plan: High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit
Updated 2026-08-15 · Risk-term list version 2026-08-15c
This is a complete, readable sample of the creative plan a product link produces. It is the whole deliverable, not an excerpt.
It contains three scored and explained ad angles, a Chinese and an English script and storyboard for each, a creator-ready brief, an A/B matrix with kill and scale criteria, a ROAS review sheet, and 5 flagged phrasings found in the seller's own copy.
We do not make the video. This document is what you hand to your creator, your production partner, or any generation tool. Demo sample, fictional product.
What this sample demonstrates
- The angles are scored: Myth-bust, Unboxing / Review, Social proof. Every point is itemised in the score breakdown inside the angle matrix below.
- The script exists in both languages, not as a translation — hooks, pacing and CTA conventions differ between them.
- The risk flags scan the seller's own copy: this sample's selling points hit 5 phrasings, each with a reason and a replacement.
- The brief is ready to send: persona, setting, props checklist, shooting and post notes, delivery specs for five placements.
- The last two sections are the ledger: the A/B matrix says what each cell tests, and the review sheet gets filled in after the flight.
This whole page was produced by the product's own pipeline and rendered by the same renderer as the "download .html" export. What you see here is what a user gets — there is no separate marketing version.
Demo sample, fictional product. The title, shop name, specs and review counts are invented and point at no real seller. The selling points deliberately keep the high-risk phrasings that real listings use, so you can see what the risk flags actually catch.
1. Product file
- TitleHigh-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit
- Chinese title高腰塑形打底裤 蹲不透 磨毛柔感
- PlatformAmazon
- Price$27.99（$39.99）
- Rating / reviews4.6 ｜ 21,740
- ShopNORWICK Studio（演示用虚构店铺）
- Data provenanceBuilt-in demo sample (hand-written, fictional product)
- URL
https://www.amazon.com/dp/B0SHAPING123/high-waist-shaping-leggings-squat-proof
Specs
| Field | Value |
|---|---|
| 面料 | 75% 锦纶 / 25% 氨纶，双面磨毛 |
| 克重 | 260g/m²（蹲姿不透） |
| 腰头 | 宽腰头 12cm，内藏隐形口袋 |
| 裤长 | 及踝 / 七分两个长度 |
| 尺码 | XS-3XL（含加长版） |
| 洗涤 | 冷水机洗，勿用柔顺剂（会破坏弹力） |
Seller’s own selling points (unedited — do not copy straight into a script)
- 260g 双面磨毛，深蹲不透光
- 宽腰头不下滑，跑步不用提裤子
- 临床验证有效矫正体态，缓解腰痛
- 穿一周小一圈，医生推荐日常穿着
- 隐形口袋能装手机，通勤和训练都能穿
2. Angle matrix
| Rank | Angle | Fit | Who | Why this angle |
|---|---|---|---|---|
| 1 | Myth-bust | 92 | Buyers who have tried this category before and were disappointed. | The category consensus is strong ("shapewear is always hot and tight"), so overturning it creates natural tension and handles the biggest objection at the same time. Product adjustments: 6 checkable specs — tests and myth-busting have material (+6); Amazon: strong search intent; tests and spec corrections work better (+6). |
| 2 | Unboxing / Review | 90 | Buyers who are comparing options and want to see the real thing before they commit. | Fabric and construction are where price scepticism lands; stretching it on camera and showing the stitching answers that directly. Product adjustments: 4.6 rating — can go on screen as evidence (+4); 6 checkable specs — tests and myth-busting have material (+6); Amazon: strong search intent; tests and spec corrections work better (+6). |
| 3 | Social proof | 89 | Buyers who are hesitating and want to know they are not the first. | Size anxiety is relieved by other people's experience — concrete body data ("I am 165cm and 60kg and wear an M") is the most effective social proof here. Product adjustments: 21,740 reviews — plenty of social-proof material (+9); 4.6 rating — can go on screen as evidence (+4). |
Score breakdown (the angle was not picked on a hunch)
- Myth-bust = 92
- Womenswear / Shapewear category prior: +80
- 6 checkable specs — tests and myth-busting have material: +6
- Amazon: strong search intent; tests and spec corrections work better: +6
- Unboxing / Review = 90
- Womenswear / Shapewear category prior: +74
- 4.6 rating — can go on screen as evidence: +4
- 6 checkable specs — tests and myth-busting have material: +6
- Amazon: strong search intent; tests and spec corrections work better: +6
- Social proof = 89
- Womenswear / Shapewear category prior: +76
- 21,740 reviews — plenty of social-proof material: +9
- 4.6 rating — can go on screen as evidence: +4
3.1 Angle: Myth-bust
Take a belief the audience already holds, show why it is wrong, and land the product as the correction.
- Fit score92
- Who it targetsBuyers who have tried this category before and were disappointed.
- Why this angleThe category consensus is strong ("shapewear is always hot and tight"), so overturning it creates natural tension and handles the biggest objection at the same time. Product adjustments: 6 checkable specs — tests and myth-busting have material (+6); Amazon: strong search intent; tests and spec corrections work better (+6).
Risk notes for this angle
- Correcting a myth still counts as a claim — do not replace one outcome promise with another.
- Naming a competitor category invites comparison rules; keep it about the belief, not a brand.
Script · Chinese (Simplified)
- 都说塑身的都闷——这件夏天我穿了一个月。
- 尺码、透气、卷边，三个问题一次说完。
- 这件我穿了一个月，说点真实的。
| Time | Beat | Voiceover | Caption |
|---|---|---|---|
| 00:00.0–00:04.8 | Hook | 都说塑身的都闷——这件夏天我穿了一个月。 | 都说塑身的都闷——这件夏天我穿了一个月 |
| 00:04.8–00:10.4 | Problem | "便宜的肯定不透气"——这句话让很多人一直没解决里面那层总往上卷。 | 所以问题一直在 |
| 00:10.4–00:18.4 | Proof | 实际情况是：勒不勒取决于压力分区做没做，不取决于价格，260g 双面磨毛，深蹲不透光。 | 实际是这样 |
| 00:18.4–00:24.8 | Demo | 所以我改成蹲下再站起来，看边缘会不会卷，薄到夏天也能穿。 | 改这一步就行 |
| 00:24.8–00:28.0 | CTA | 链接在下面，拿不准尺码就问我 | 链接在下面，拿不准尺码就问我 |
Script · English (US)
- Everyone says these are hot to wear. I wore this all summer.
- Sizing, breathability, rolling — all three, in one video.
- I have worn this for a month. Here is the honest version.
| Time | Beat | Voiceover | Caption |
|---|---|---|---|
| 00:00.0–00:04.8 | Hook | Everyone says these are hot to wear. I wore this all summer. | I wore this all summer |
| 00:04.8–00:10.4 | Problem | "a cheap one cannot possibly breathe" — that one belief is why the layer underneath keeps rolling up never gets fixed. | Why it never gets fixed |
| 00:10.4–00:18.4 | Proof | Here is what is actually true: whether it digs in comes down to pressure zoning, not price. knit in one piece, laser-cut edges, no stitched seam. | What is actually true |
| 00:18.4–00:24.8 | Demo | So I switched to squat down and stand back up, watch the edge, and you cannot tell there is a layer underneath. | Change this one step |
| 00:24.8–00:28.0 | CTA | Tap the cart, they sell out fast | Tap the cart, they sell out fast |
Storyboard · voiceover language Chinese (Simplified)
| # | Time | Visual | Camera | Voiceover | Caption | Props | Music |
|---|---|---|---|---|---|---|---|
| 1 | 00:00.0–00:04.8 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | This shot decides completion: action first, core message inside 3 seconds | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Angle reminder: Correcting a myth still counts as a claim — do not replace one outcome promise with another. | 都说塑身的都闷——这件夏天我穿了一个月。 | 都说塑身的都闷——这件夏天我穿了一个月 | A fitted dress (used for the comparison, never changed) | No music or ambient only — the voice carries it; any bed in the first 3 seconds dilutes the message |
| 2 | 00:04.8–00:10.4 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Shoot "life without it" — the more specific the better, no staged feel | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | "便宜的肯定不透气"——这句话让很多人一直没解决里面那层总往上卷。 | 所以问题一直在 | Close-up of the size label | Low-frequency bed at -24 dB, a touch of tension |
| 3 | 00:10.4–00:18.4 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot something checkable (a spec, a timer, a seam) and overlay the number | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | 实际情况是：勒不勒取决于压力分区做没做，不取决于价格，260g 双面磨毛，深蹲不透光。 | 实际是这样 | An older piece of the same type (the control) | Duck the bed to -26 dB so the numbers are clear |
| 4 | 00:18.4–00:24.8 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot the whole action uncut; the viewer has to see how it is done | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | 所以我改成蹲下再站起来，看边缘会不会卷，薄到夏天也能穿。 | 改这一步就行 | A fitting mirror | Upbeat bed comes in at -18 dB, cut to the action |
| 5 | 00:24.8–00:28.0 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Look at the lens, bring the tone down, hold 1 s at the end | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | 链接在下面，拿不准尺码就问我 | 链接在下面，拿不准尺码就问我 | A fitted dress (used for the comparison, never changed) | Outro sting + 1 s hold so it loops cleanly |
Storyboard · voiceover language English (US)
| # | Time | Visual | Camera | Voiceover | Caption | Props | Music |
|---|---|---|---|---|---|---|---|
| 1 | 00:00.0–00:04.8 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | This shot decides completion: action first, core message inside 3 seconds | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Angle reminder: Correcting a myth still counts as a claim — do not replace one outcome promise with another. | Everyone says these are hot to wear. I wore this all summer. | I wore this all summer | A fitted dress (used for the comparison, never changed) | No music or ambient only — the voice carries it; any bed in the first 3 seconds dilutes the message |
| 2 | 00:04.8–00:10.4 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Shoot "life without it" — the more specific the better, no staged feel | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | "a cheap one cannot possibly breathe" — that one belief is why the layer underneath keeps rolling up never gets fixed. | Why it never gets fixed | Close-up of the size label | Low-frequency bed at -24 dB, a touch of tension |
| 3 | 00:10.4–00:18.4 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot something checkable (a spec, a timer, a seam) and overlay the number | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | Here is what is actually true: whether it digs in comes down to pressure zoning, not price. knit in one piece, laser-cut edges, no stitched seam. | What is actually true | An older piece of the same type (the control) | Duck the bed to -26 dB so the numbers are clear |
| 4 | 00:18.4–00:24.8 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot the whole action uncut; the viewer has to see how it is done | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | So I switched to squat down and stand back up, watch the edge, and you cannot tell there is a layer underneath. | Change this one step | A fitting mirror | Upbeat bed comes in at -18 dB, cut to the action |
| 5 | 00:24.8–00:28.0 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Look at the lens, bring the tone down, hold 1 s at the end | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Tap the cart, they sell out fast | Tap the cart, they sell out fast | A fitted dress (used for the comparison, never changed) | Outro sting + 1 s hold so it loops cleanly |
Creative brief (send this to a creator or production partner as-is)
- On-camera personaA woman aged 24–35 with a real body (not a model figure) who can show a before-and-after naturally (24-35)
- StylingBare-faced to light makeup, two outer looks (one fitted dress, one workwear set)
- ToneCandid, sharing-from-experience — no "perfect body" narrative
- SettingBedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim
Props checklist
- A fitted dress (used for the comparison, never changed)
- Close-up of the size label
- An older piece of the same type (the control)
- A fitting mirror
- The product itself: High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit
- Price card asset (list price and current price on their own layer, so the price can be swapped)
Shooting notes
- A before-and-after must change exactly one variable — same outer layer, same camera position, same lighting, or it is not credible
- "No roll after sitting all day" needs timestamp evidence: shoot once at the start and once at the end
- Keep scales, tape measures and body numbers out of frame entirely — they trigger health-category review directly
- Narrative premise of this angle: Take a belief the audience already holds, show why it is wrong, and land the product as the correction.
- Risk reminder: Correcting a myth still counts as a claim — do not replace one outcome promise with another.
- Risk reminder: Naming a competitor category invites comparison rules; keep it about the belief, not a brand.
- Phrasings flagged for review — keep them out of the voiceover and captions:
矫正体态,缓解腰痛,临床验证,小一圈,医生推荐(reasons and alternatives are in the "Risk flags / items to check" section)
Post notes
- Show the comparison as picture-in-picture or a slider split, not a jump cut
- Captions must follow the cleaned copy exactly; the voiceover can be looser but a flagged term must never appear in the captions
- End on a sizing note — it reduces returns
- [Chinese (Simplified)] Fast delivery and high information density; no pleasantries in the first 3 seconds
- [Chinese (Simplified)] Keep the listing currency ($). If this version is ever run in a Chinese-language market, swap in that market’s currency and real price — never convert out loud
- [Chinese (Simplified)] No absolutes (best / #1 / permanent): China’s Advertising Law fines them directly
- [English (US)] First person, conversational; pauses and self-deprecation are fine — sounding scripted kills conversion
- [English (US)] Use concrete numbers ("3 weeks", not "a long time"); they read as more credible
- [English (US)] The FTC requires a clear disclosure of the commercial relationship (#ad / Paid partnership)
- [Chinese (Simplified)] Must appear: 本品为服饰，不具备减重、医疗或矫正功能
- [Chinese (Simplified)] Must appear: 视觉效果因个人身材与所选尺码而异
- [English (US)] Must appear: This is a garment. It does not provide weight-loss, medical, or corrective effects.
- [English (US)] Must appear: How it looks varies by body and by the size you pick.
Delivery specs (one shoot, two aspect ratios)
| Platform | Aspect | Max length | Safe area | Notes |
|---|---|---|---|---|
| TikTok / TikTok Shop | 9:16 | 60s | Leave roughly 130px clear at the top and 480px at the bottom; the interaction rail on the right takes about 140px | Land the core message in the first 3 seconds; captions are on by default; AI-generated content must be disclosed by you, and the platform also labels it automatically |
| Meta (Reels and Feed) | 9:16 as the main cut, with a 1:1 delivered as backup | 30s | Keep text, logo and key elements out of roughly the top 14%, the bottom 35% and 6% on each side | Advantage+ crops automatically, so keep key elements centred; also deliver a version that reads with the sound off |
| Temu on-platform | 1:1 and 3:4 | 30s | About 8% margin on all four sides | The product must appear within the first 2 seconds and fill at least 40% of the frame; do not burn price tags into the footage, so the cut survives a price change |
| Amazon Sponsored Brands | 16:9 — the spec page says "only"; there is no vertical option | 45s | Keep text and key elements out of the bottom-right corner; no black, colour or blurred bars | 6–45 seconds, and the spec page strongly recommends 20 or under; the video language must match the marketplace (an English marketplace needs English captions or an English voiceover) |
| YouTube Shorts | 9:16 | 60s | Keep the bottom 15% clear of the title bar | Watch-through carries most of the weight; hold a still final frame for about a second so the loop feels intentional |
Risk flags / items to check
These risk flags are an assisted check. They are not a compliance ruling and not legal advice — the platform’s review and your own legal judgment decide.
| Term | Level | List | Found in | Why it is risky | Say instead |
|---|---|---|---|---|---|
矫正体态 | High risk | Chinese list | Seller's listing copy | A therapeutic claim. "Gives your back something to lean into" is a sensation; "corrects posture" is a treatment outcome. | gives your back something to lean into |
缓解腰痛 | High risk | Chinese list | Seller's listing copy | An explicit symptom-treatment claim; it moves the listing into health review. | (no safe alternative — cut the whole sentence) |
临床验证 | High risk | Chinese list | Seller's listing copy | An evidence claim requiring a study you can show. There is almost no published work that supports this sentence for apparel. | (no safe alternative — cut the whole sentence) |
小一圈 | High risk | Chinese list | Seller's listing copy | A quantified body-change promise — the highest-risk phrasing in this list because it can be checked line by line. | looks smoother under a fitted dress |
医生推荐 | High risk | Chinese list | Seller's listing copy | An endorsement claim. You have to be able to produce the doctor; without one it is a factual statement you can be held to. | (no safe alternative — cut the whole sentence) |
- 本品为服饰，不具备减重、医疗或矫正功能
- 视觉效果因个人身材与所选尺码而异
| Term | Level | List | Found in | Why it is risky | Say instead |
|---|---|---|---|---|---|
矫正体态 | High risk | Chinese list | Seller's listing copy | A therapeutic claim. "Gives your back something to lean into" is a sensation; "corrects posture" is a treatment outcome. | gives your back something to lean into |
缓解腰痛 | High risk | Chinese list | Seller's listing copy | An explicit symptom-treatment claim; it moves the listing into health review. | (no safe alternative — cut the whole sentence) |
临床验证 | High risk | Chinese list | Seller's listing copy | An evidence claim requiring a study you can show. There is almost no published work that supports this sentence for apparel. | (no safe alternative — cut the whole sentence) |
小一圈 | High risk | Chinese list | Seller's listing copy | A quantified body-change promise — the highest-risk phrasing in this list because it can be checked line by line. | looks smoother under a fitted dress |
医生推荐 | High risk | Chinese list | Seller's listing copy | An endorsement claim. You have to be able to produce the doctor; without one it is a factual statement you can be held to. | (no safe alternative — cut the whole sentence) |
- This is a garment. It does not provide weight-loss, medical, or corrective effects.
- How it looks varies by body and by the size you pick.
3.2 Angle: Unboxing / Review
Use the credibility of a first-person, hands-on review to carry the conversion.
- Fit score90
- Who it targetsBuyers who are comparing options and want to see the real thing before they commit.
- Why this angleFabric and construction are where price scepticism lands; stretching it on camera and showing the stitching answers that directly. Product adjustments: 4.6 rating — can go on screen as evidence (+4); 6 checkable specs — tests and myth-busting have material (+6); Amazon: strong search intent; tests and spec corrections work better (+6).
Risk notes for this angle
- A first-person account ("I wore this all day") is far safer than a second-person claim ("this will slim you").
- If a creator ad-libs an outcome promise, the platform judges the published video, not who wrote the line.
Script · Chinese (Simplified)
- 这件我坐了一整天，没卷边。
- 打底裤我只测一件事：蹲下去透不透。
- 尺码、透气、卷边，三个问题一次说完。
| Time | Beat | Voiceover | Caption |
|---|---|---|---|
| 00:00.0–00:04.0 | Hook | 这件我坐了一整天，没卷边。 | 这件我坐了一整天，没卷边 |
| 00:04.0–00:13.6 | Demo | 直接上手测：蹲下再站起来，看边缘会不会卷，全程没有剪辑。 | 实测，无剪辑 |
| 00:13.6–00:20.8 | Proof | 78% 锦纶 22% 氨纶，四面弹——这是我最在意的一项，它过了。 | 78% 锦纶 22% 氨纶，四面弹 |
| 00:20.8–00:24.8 | Offer | 现在 $27.99，这个价位这个表现，我觉得说得过去。 | 现在 $27.99 |
| 00:24.8–00:28.0 | CTA | 想要的点小黄车，库存不多 | 想要的点小黄车，库存不多 |
Script · English (US)
- I sat in this for a full day. It did not roll.
- I test leggings on one thing: whether they go see-through when you squat.
- Sizing, breathability, rolling — all three, in one video.
| Time | Beat | Voiceover | Caption |
|---|---|---|---|
| 00:00.0–00:04.0 | Hook | I sat in this for a full day. It did not roll. | I sat in this for a full day |
| 00:04.0–00:13.6 | Demo | Testing it right now: squat down and stand back up, watch the edge. No cuts, one take. | One take, no cuts |
| 00:13.6–00:20.8 | Proof | 78% nylon, 22% spandex, four-way stretch — that was the one thing I cared about, and it passed. | 78% nylon, 22% spandex, four-way stretch |
| 00:20.8–00:24.8 | Offer | it is $27.99 right now. For this price and this result, that math works. | it is $27.99 right now |
| 00:24.8–00:28.0 | CTA | Link is in the bio — coupon applies at checkout | Link is in the bio — coupon applies at checkout |
Storyboard · voiceover language Chinese (Simplified)
| # | Time | Visual | Camera | Voiceover | Caption | Props | Music |
|---|---|---|---|---|---|---|---|
| 1 | 00:00.0–00:04.0 | Handheld vertical (slight shake is fine), Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; on camera: A woman aged 24–35 with a real body (not a model figure) who can show a before-and-after naturally. Action first — pull the dress over it and see if you can tell | This shot decides completion: action first, core message inside 3 seconds | Handheld at chest height, slow push-in; do not over-stabilise — the "real" look comes first | Angle reminder: A first-person account ("I wore this all day") is far safer than a second-person claim ("this will slim you"). | 这件我坐了一整天，没卷边。 | 这件我坐了一整天，没卷边 | A fitted dress (used for the comparison, never changed) | No music or ambient only — the voice carries it; any bed in the first 3 seconds dilutes the message |
| 2 | 00:04.0–00:13.6 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot the whole action uncut; the viewer has to see how it is done | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | 直接上手测：蹲下再站起来，看边缘会不会卷，全程没有剪辑。 | 实测，无剪辑 | Close-up of the size label | Upbeat bed comes in at -18 dB, cut to the action |
| 3 | 00:13.6–00:20.8 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot something checkable (a spec, a timer, a seam) and overlay the number | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | 78% 锦纶 22% 氨纶，四面弹——这是我最在意的一项，它过了。 | 78% 锦纶 22% 氨纶，四面弹 | An older piece of the same type (the control) | Duck the bed to -26 dB so the numbers are clear |
| 4 | 00:20.8–00:24.8 | Full-screen text card over a product still (High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit); the text must match the voiceover so the price can be swapped later | Price and offer on their own layer so they can be swapped — do not burn them in | Still frame, composited in the editing software — nothing to shoot | 现在 $27.99，这个价位这个表现，我觉得说得过去。 | 现在 $27.99 | A fitting mirror | Energy rises, one short sting |
| 5 | 00:24.8–00:28.0 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Look at the lens, bring the tone down, hold 1 s at the end | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | 想要的点小黄车，库存不多 | 想要的点小黄车，库存不多 | A fitted dress (used for the comparison, never changed) | Outro sting + 1 s hold so it loops cleanly |
Storyboard · voiceover language English (US)
| # | Time | Visual | Camera | Voiceover | Caption | Props | Music |
|---|---|---|---|---|---|---|---|
| 1 | 00:00.0–00:04.0 | Handheld vertical (slight shake is fine), Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; on camera: A woman aged 24–35 with a real body (not a model figure) who can show a before-and-after naturally. Action first — pull the dress over it and see if you can tell | This shot decides completion: action first, core message inside 3 seconds | Handheld at chest height, slow push-in; do not over-stabilise — the "real" look comes first | Angle reminder: A first-person account ("I wore this all day") is far safer than a second-person claim ("this will slim you"). | I sat in this for a full day. It did not roll. | I sat in this for a full day | A fitted dress (used for the comparison, never changed) | No music or ambient only — the voice carries it; any bed in the first 3 seconds dilutes the message |
| 2 | 00:04.0–00:13.6 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot the whole action uncut; the viewer has to see how it is done | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | Testing it right now: squat down and stand back up, watch the edge. No cuts, one take. | One take, no cuts | Close-up of the size label | Upbeat bed comes in at -18 dB, cut to the action |
| 3 | 00:13.6–00:20.8 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot something checkable (a spec, a timer, a seam) and overlay the number | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | 78% nylon, 22% spandex, four-way stretch — that was the one thing I cared about, and it passed. | 78% nylon, 22% spandex, four-way stretch | An older piece of the same type (the control) | Duck the bed to -26 dB so the numbers are clear |
| 4 | 00:20.8–00:24.8 | Full-screen text card over a product still (High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit); the text must match the voiceover so the price can be swapped later | Price and offer on their own layer so they can be swapped — do not burn them in | Still frame, composited in the editing software — nothing to shoot | it is $27.99 right now. For this price and this result, that math works. | it is $27.99 right now | A fitting mirror | Energy rises, one short sting |
| 5 | 00:24.8–00:28.0 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Look at the lens, bring the tone down, hold 1 s at the end | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Link is in the bio — coupon applies at checkout | Link is in the bio — coupon applies at checkout | A fitted dress (used for the comparison, never changed) | Outro sting + 1 s hold so it loops cleanly |
Creative brief (send this to a creator or production partner as-is)
- On-camera personaA woman aged 24–35 with a real body (not a model figure) who can show a before-and-after naturally (24-35)
- StylingBare-faced to light makeup, two outer looks (one fitted dress, one workwear set)
- ToneCandid, sharing-from-experience — no "perfect body" narrative
- SettingBedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim
Props checklist
- A fitted dress (used for the comparison, never changed)
- Close-up of the size label
- An older piece of the same type (the control)
- A fitting mirror
- The product itself: High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit
- Price card asset (list price and current price on their own layer, so the price can be swapped)
Shooting notes
- A before-and-after must change exactly one variable — same outer layer, same camera position, same lighting, or it is not credible
- "No roll after sitting all day" needs timestamp evidence: shoot once at the start and once at the end
- Keep scales, tape measures and body numbers out of frame entirely — they trigger health-category review directly
- Narrative premise of this angle: Use the credibility of a first-person, hands-on review to carry the conversion.
- Risk reminder: A first-person account ("I wore this all day") is far safer than a second-person claim ("this will slim you").
- Risk reminder: If a creator ad-libs an outcome promise, the platform judges the published video, not who wrote the line.
- Phrasings flagged for review — keep them out of the voiceover and captions:
矫正体态,缓解腰痛,临床验证,小一圈,医生推荐(reasons and alternatives are in the "Risk flags / items to check" section)
Post notes
- Show the comparison as picture-in-picture or a slider split, not a jump cut
- Captions must follow the cleaned copy exactly; the voiceover can be looser but a flagged term must never appear in the captions
- End on a sizing note — it reduces returns
- [Chinese (Simplified)] Fast delivery and high information density; no pleasantries in the first 3 seconds
- [Chinese (Simplified)] Keep the listing currency ($). If this version is ever run in a Chinese-language market, swap in that market’s currency and real price — never convert out loud
- [Chinese (Simplified)] No absolutes (best / #1 / permanent): China’s Advertising Law fines them directly
- [English (US)] First person, conversational; pauses and self-deprecation are fine — sounding scripted kills conversion
- [English (US)] Use concrete numbers ("3 weeks", not "a long time"); they read as more credible
- [English (US)] The FTC requires a clear disclosure of the commercial relationship (#ad / Paid partnership)
- [Chinese (Simplified)] Must appear: 本品为服饰，不具备减重、医疗或矫正功能
- [Chinese (Simplified)] Must appear: 视觉效果因个人身材与所选尺码而异
- [English (US)] Must appear: This is a garment. It does not provide weight-loss, medical, or corrective effects.
- [English (US)] Must appear: How it looks varies by body and by the size you pick.
Delivery specs (one shoot, two aspect ratios)
| Platform | Aspect | Max length | Safe area | Notes |
|---|---|---|---|---|
| TikTok / TikTok Shop | 9:16 | 60s | Leave roughly 130px clear at the top and 480px at the bottom; the interaction rail on the right takes about 140px | Land the core message in the first 3 seconds; captions are on by default; AI-generated content must be disclosed by you, and the platform also labels it automatically |
| Meta (Reels and Feed) | 9:16 as the main cut, with a 1:1 delivered as backup | 30s | Keep text, logo and key elements out of roughly the top 14%, the bottom 35% and 6% on each side | Advantage+ crops automatically, so keep key elements centred; also deliver a version that reads with the sound off |
| Temu on-platform | 1:1 and 3:4 | 30s | About 8% margin on all four sides | The product must appear within the first 2 seconds and fill at least 40% of the frame; do not burn price tags into the footage, so the cut survives a price change |
| Amazon Sponsored Brands | 16:9 — the spec page says "only"; there is no vertical option | 45s | Keep text and key elements out of the bottom-right corner; no black, colour or blurred bars | 6–45 seconds, and the spec page strongly recommends 20 or under; the video language must match the marketplace (an English marketplace needs English captions or an English voiceover) |
| YouTube Shorts | 9:16 | 60s | Keep the bottom 15% clear of the title bar | Watch-through carries most of the weight; hold a still final frame for about a second so the loop feels intentional |
Risk flags / items to check
These risk flags are an assisted check. They are not a compliance ruling and not legal advice — the platform’s review and your own legal judgment decide.
| Term | Level | List | Found in | Why it is risky | Say instead |
|---|---|---|---|---|---|
矫正体态 | High risk | Chinese list | Seller's listing copy | A therapeutic claim. "Gives your back something to lean into" is a sensation; "corrects posture" is a treatment outcome. | gives your back something to lean into |
缓解腰痛 | High risk | Chinese list | Seller's listing copy | An explicit symptom-treatment claim; it moves the listing into health review. | (no safe alternative — cut the whole sentence) |
临床验证 | High risk | Chinese list | Seller's listing copy | An evidence claim requiring a study you can show. There is almost no published work that supports this sentence for apparel. | (no safe alternative — cut the whole sentence) |
小一圈 | High risk | Chinese list | Seller's listing copy | A quantified body-change promise — the highest-risk phrasing in this list because it can be checked line by line. | looks smoother under a fitted dress |
医生推荐 | High risk | Chinese list | Seller's listing copy | An endorsement claim. You have to be able to produce the doctor; without one it is a factual statement you can be held to. | (no safe alternative — cut the whole sentence) |
- 本品为服饰，不具备减重、医疗或矫正功能
- 视觉效果因个人身材与所选尺码而异
| Term | Level | List | Found in | Why it is risky | Say instead |
|---|---|---|---|---|---|
矫正体态 | High risk | Chinese list | Seller's listing copy | A therapeutic claim. "Gives your back something to lean into" is a sensation; "corrects posture" is a treatment outcome. | gives your back something to lean into |
缓解腰痛 | High risk | Chinese list | Seller's listing copy | An explicit symptom-treatment claim; it moves the listing into health review. | (no safe alternative — cut the whole sentence) |
临床验证 | High risk | Chinese list | Seller's listing copy | An evidence claim requiring a study you can show. There is almost no published work that supports this sentence for apparel. | (no safe alternative — cut the whole sentence) |
小一圈 | High risk | Chinese list | Seller's listing copy | A quantified body-change promise — the highest-risk phrasing in this list because it can be checked line by line. | looks smoother under a fitted dress |
医生推荐 | High risk | Chinese list | Seller's listing copy | An endorsement claim. You have to be able to produce the doctor; without one it is a factual statement you can be held to. | (no safe alternative — cut the whole sentence) |
- This is a garment. It does not provide weight-loss, medical, or corrective effects.
- How it looks varies by body and by the size you pick.
3.3 Angle: Social proof
Lead with the fact that many people already bought it and were happy.
- Fit score89
- Who it targetsBuyers who are hesitating and want to know they are not the first.
- Why this angleSize anxiety is relieved by other people's experience — concrete body data ("I am 165cm and 60kg and wear an M") is the most effective social proof here. Product adjustments: 21,740 reviews — plenty of social-proof material (+9); 4.6 rating — can go on screen as evidence (+4).
Risk notes for this angle
- Review counts and ratings must be real and current; a stale number is a false claim.
- Do not stage testimonials or imply an endorsement you cannot produce.
Script · Chinese (Simplified)
- 21,740 条评价、4.6 分，尺码怎么选我一次说清楚。
- 这件我穿了一个月，说点真实的。
- 尺码、透气、卷边，三个问题一次说完。
| Time | Beat | Voiceover | Caption |
|---|---|---|---|
| 00:00.0–00:04.2 | Hook | 21,740 条评价、4.6 分，尺码怎么选我一次说清楚。 | 尺码怎么选我一次说清楚 |
| 00:04.2–00:12.7 | Proof | 评价里被提到最多的一句是"外穿看不出里面有一层"，评价里最多的一句是"外穿真的看不出来"。 | 评价里提最多的一点 |
| 00:12.7–00:20.3 | Demo | 我拿到手也确认了：把裙子套上去，看外面看不看得出来，薄到夏天也能穿。 | 我也验证了 |
| 00:20.3–00:24.5 | Offer | $27.99，两件的价格还不到一件大牌的 | $27.99，两件的价格还不到一件大牌的 |
| 00:24.5–00:27.9 | CTA | 想要的点小黄车，库存不多 | 想要的点小黄车，库存不多 |
Script · English (US)
- 21,740 reviews at 4.6 stars — let me settle the sizing question once.
- I have worn this for a month. Here is the honest version.
- Sizing, breathability, rolling — all three, in one video.
| Time | Beat | Voiceover | Caption |
|---|---|---|---|
| 00:00.0–00:04.2 | Hook | 21,740 reviews at 4.6 stars — let me settle the sizing question once. | 21,740 reviews at 4.6 stars — let me settle the sizing question once |
| 00:04.2–00:12.7 | Proof | The line that shows up most in the reviews is "it is thin enough to wear in summer" — the most repeated review line is "you really cannot see it". | Most-repeated review line |
| 00:12.7–00:20.3 | Demo | I checked it myself too: pull the dress over it and see if you can tell, you cannot tell there is a layer underneath. | Verified it myself |
| 00:20.3–00:24.5 | Offer | $27.99 — two of these still cost less than one name-brand | $27.99 — two of these still cost less than one name-brand |
| 00:24.5–00:27.9 | CTA | Link in bio — ask me if you are between sizes | Link in bio — ask me if you are between sizes |
Storyboard · voiceover language Chinese (Simplified)
| # | Time | Visual | Camera | Voiceover | Caption | Props | Music |
|---|---|---|---|---|---|---|---|
| 1 | 00:00.0–00:04.2 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | This shot decides completion: action first, core message inside 3 seconds | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Angle reminder: Review counts and ratings must be real and current; a stale number is a false claim. | 21,740 条评价、4.6 分，尺码怎么选我一次说清楚。 | 尺码怎么选我一次说清楚 | A fitted dress (used for the comparison, never changed) | No music or ambient only — the voice carries it; any bed in the first 3 seconds dilutes the message |
| 2 | 00:04.2–00:12.7 | Full-screen text card over a product still (High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit); the text must match the voiceover so the price can be swapped later | Shoot something checkable (a spec, a timer, a seam) and overlay the number | Still frame, composited in the editing software — nothing to shoot | 评价里被提到最多的一句是"外穿看不出里面有一层"，评价里最多的一句是"外穿真的看不出来"。 | 评价里提最多的一点 | Close-up of the size label | Duck the bed to -26 dB so the numbers are clear |
| 3 | 00:12.7–00:20.3 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot the whole action uncut; the viewer has to see how it is done | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | 我拿到手也确认了：把裙子套上去，看外面看不看得出来，薄到夏天也能穿。 | 我也验证了 | An older piece of the same type (the control) | Upbeat bed comes in at -18 dB, cut to the action |
| 4 | 00:20.3–00:24.5 | Full-screen text card over a product still (High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit); the text must match the voiceover so the price can be swapped later | Price and offer on their own layer so they can be swapped — do not burn them in | Still frame, composited in the editing software — nothing to shoot | $27.99，两件的价格还不到一件大牌的 | $27.99，两件的价格还不到一件大牌的 | A fitting mirror | Energy rises, one short sting |
| 5 | 00:24.5–00:27.9 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Look at the lens, bring the tone down, hold 1 s at the end | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | 想要的点小黄车，库存不多 | 想要的点小黄车，库存不多 | A fitted dress (used for the comparison, never changed) | Outro sting + 1 s hold so it loops cleanly |
Storyboard · voiceover language English (US)
| # | Time | Visual | Camera | Voiceover | Caption | Props | Music |
|---|---|---|---|---|---|---|---|
| 1 | 00:00.0–00:04.2 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | This shot decides completion: action first, core message inside 3 seconds | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Angle reminder: Review counts and ratings must be real and current; a stale number is a false claim. | 21,740 reviews at 4.6 stars — let me settle the sizing question once. | 21,740 reviews at 4.6 stars — let me settle the sizing question once | A fitted dress (used for the comparison, never changed) | No music or ambient only — the voice carries it; any bed in the first 3 seconds dilutes the message |
| 2 | 00:04.2–00:12.7 | Full-screen text card over a product still (High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit); the text must match the voiceover so the price can be swapped later | Shoot something checkable (a spec, a timer, a seam) and overlay the number | Still frame, composited in the editing software — nothing to shoot | The line that shows up most in the reviews is "it is thin enough to wear in summer" — the most repeated review line is "you really cannot see it". | Most-repeated review line | Close-up of the size label | Duck the bed to -26 dB so the numbers are clear |
| 3 | 00:12.7–00:20.3 | Product close-up / macro, shallow depth of field, focus on High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit; keep A fitted dress (used for the comparison, never changed) in frame for scale | Shoot the whole action uncut; the viewer has to see how it is done | Locked-off macro, f/2.0–2.8, manual focus locked — no autofocus hunting | I checked it myself too: pull the dress over it and see if you can tell, you cannot tell there is a layer underneath. | Verified it myself | An older piece of the same type (the control) | Upbeat bed comes in at -18 dB, cut to the action |
| 4 | 00:20.3–00:24.5 | Full-screen text card over a product still (High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit); the text must match the voiceover so the price can be swapped later | Price and offer on their own layer so they can be swapped — do not burn them in | Still frame, composited in the editing software — nothing to shoot | $27.99 — two of these still cost less than one name-brand | $27.99 — two of these still cost less than one name-brand | A fitting mirror | Energy rises, one short sting |
| 5 | 00:24.5–00:27.9 | Waist-up talking head, eyes on the lens, background: Bedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim; no teleprompter feel | Look at the lens, bring the tone down, hold 1 s at the end | Locked-off, 35mm equivalent, subject centred slightly high, leave the bottom clear for captions | Link in bio — ask me if you are between sizes | Link in bio — ask me if you are between sizes | A fitted dress (used for the comparison, never changed) | Outro sting + 1 s hold so it loops cleanly |
Creative brief (send this to a creator or production partner as-is)
- On-camera personaA woman aged 24–35 with a real body (not a model figure) who can show a before-and-after naturally (24-35)
- StylingBare-faced to light makeup, two outer looks (one fitted dress, one workwear set)
- ToneCandid, sharing-from-experience — no "perfect body" narrative
- SettingBedroom or in front of a fitting mirror in natural light, plus a real outdoor or office segment as evidence for the "wore it all day" claim
Props checklist
- A fitted dress (used for the comparison, never changed)
- Close-up of the size label
- An older piece of the same type (the control)
- A fitting mirror
- The product itself: High-Waist Shaping Leggings, Squat-Proof, Buttery-Soft Brushed Knit
- Price card asset (list price and current price on their own layer, so the price can be swapped)
Shooting notes
- A before-and-after must change exactly one variable — same outer layer, same camera position, same lighting, or it is not credible
- "No roll after sitting all day" needs timestamp evidence: shoot once at the start and once at the end
- Keep scales, tape measures and body numbers out of frame entirely — they trigger health-category review directly
- Narrative premise of this angle: Lead with the fact that many people already bought it and were happy.
- Risk reminder: Review counts and ratings must be real and current; a stale number is a false claim.
- Risk reminder: Do not stage testimonials or imply an endorsement you cannot produce.
- Phrasings flagged for review — keep them out of the voiceover and captions:
矫正体态,缓解腰痛,临床验证,小一圈,医生推荐(reasons and alternatives are in the "Risk flags / items to check" section)
Post notes
- Show the comparison as picture-in-picture or a slider split, not a jump cut
- Captions must follow the cleaned copy exactly; the voiceover can be looser but a flagged term must never appear in the captions
- End on a sizing note — it reduces returns
- [Chinese (Simplified)] Fast delivery and high information density; no pleasantries in the first 3 seconds
- [Chinese (Simplified)] Keep the listing currency ($). If this version is ever run in a Chinese-language market, swap in that market’s currency and real price — never convert out loud
- [Chinese (Simplified)] No absolutes (best / #1 / permanent): China’s Advertising Law fines them directly
- [English (US)] First person, conversational; pauses and self-deprecation are fine — sounding scripted kills conversion
- [English (US)] Use concrete numbers ("3 weeks", not "a long time"); they read as more credible
- [English (US)] The FTC requires a clear disclosure of the commercial relationship (#ad / Paid partnership)
- [Chinese (Simplified)] Must appear: 本品为服饰，不具备减重、医疗或矫正功能
- [Chinese (Simplified)] Must appear: 视觉效果因个人身材与所选尺码而异
- [English (US)] Must appear: This is a garment. It does not provide weight-loss, medical, or corrective effects.
- [English (US)] Must appear: How it looks varies by body and by the size you pick.
Delivery specs (one shoot, two aspect ratios)
| Platform | Aspect | Max length | Safe area | Notes |
|---|---|---|---|---|
| TikTok / TikTok Shop | 9:16 | 60s | Leave roughly 130px clear at the top and 480px at the bottom; the interaction rail on the right takes about 140px | Land the core message in the first 3 seconds; captions are on by default; AI-generated content must be disclosed by you, and the platform also labels it automatically |
| Meta (Reels and Feed) | 9:16 as the main cut, with a 1:1 delivered as backup | 30s | Keep text, logo and key elements out of roughly the top 14%, the bottom 35% and 6% on each side | Advantage+ crops automatically, so keep key elements centred; also deliver a version that reads with the sound off |
| Temu on-platform | 1:1 and 3:4 | 30s | About 8% margin on all four sides | The product must appear within the first 2 seconds and fill at least 40% of the frame; do not burn price tags into the footage, so the cut survives a price change |
| Amazon Sponsored Brands | 16:9 — the spec page says "only"; there is no vertical option | 45s | Keep text and key elements out of the bottom-right corner; no black, colour or blurred bars | 6–45 seconds, and the spec page strongly recommends 20 or under; the video language must match the marketplace (an English marketplace needs English captions or an English voiceover) |
| YouTube Shorts | 9:16 | 60s | Keep the bottom 15% clear of the title bar | Watch-through carries most of the weight; hold a still final frame for about a second so the loop feels intentional |
Risk flags / items to check
These risk flags are an assisted check. They are not a compliance ruling and not legal advice — the platform’s review and your own legal judgment decide.
| Term | Level | List | Found in | Why it is risky | Say instead |
|---|---|---|---|---|---|
矫正体态 | High risk | Chinese list | Seller's listing copy | A therapeutic claim. "Gives your back something to lean into" is a sensation; "corrects posture" is a treatment outcome. | gives your back something to lean into |
缓解腰痛 | High risk | Chinese list | Seller's listing copy | An explicit symptom-treatment claim; it moves the listing into health review. | (no safe alternative — cut the whole sentence) |
临床验证 | High risk | Chinese list | Seller's listing copy | An evidence claim requiring a study you can show. There is almost no published work that supports this sentence for apparel. | (no safe alternative — cut the whole sentence) |
小一圈 | High risk | Chinese list | Seller's listing copy | A quantified body-change promise — the highest-risk phrasing in this list because it can be checked line by line. | looks smoother under a fitted dress |
医生推荐 | High risk | Chinese list | Seller's listing copy | An endorsement claim. You have to be able to produce the doctor; without one it is a factual statement you can be held to. | (no safe alternative — cut the whole sentence) |
- 本品为服饰，不具备减重、医疗或矫正功能
- 视觉效果因个人身材与所选尺码而异
| Term | Level | List | Found in | Why it is risky | Say instead |
|---|---|---|---|---|---|
矫正体态 | High risk | Chinese list | Seller's listing copy | A therapeutic claim. "Gives your back something to lean into" is a sensation; "corrects posture" is a treatment outcome. | gives your back something to lean into |
缓解腰痛 | High risk | Chinese list | Seller's listing copy | An explicit symptom-treatment claim; it moves the listing into health review. | (no safe alternative — cut the whole sentence) |
临床验证 | High risk | Chinese list | Seller's listing copy | An evidence claim requiring a study you can show. There is almost no published work that supports this sentence for apparel. | (no safe alternative — cut the whole sentence) |
小一圈 | High risk | Chinese list | Seller's listing copy | A quantified body-change promise — the highest-risk phrasing in this list because it can be checked line by line. | looks smoother under a fitted dress |
医生推荐 | High risk | Chinese list | Seller's listing copy | An endorsement claim. You have to be able to produce the doctor; without one it is a factual statement you can be held to. | (no safe alternative — cut the whole sentence) |
- This is a garment. It does not provide weight-loss, medical, or corrective effects.
- How it looks varies by body and by the size you pick.
4. A/B test matrix
| Cell | Angle | Hook | Budget share | What this cell tests |
|---|---|---|---|---|
| A | Myth-bust | Everyone says these are hot to wear. I wore this all summer. | 21.2% | Tests whether the "Myth-bust" angle itself holds in this category / market |
| B | Myth-bust | Sizing, breathability, rolling — all three, in one video. | 12.7% | Same angle, different hook — separates "the angle failed" from "the hook failed" so the angle is not killed by mistake |
| C | Unboxing / Review | I sat in this for a full day. It did not roll. | 20.8% | Tests whether the "Unboxing / Review" angle itself holds in this category / market |
| D | Unboxing / Review | I test leggings on one thing: whether they go see-through when you squat. | 12.5% | Same angle, different hook — separates "the angle failed" from "the hook failed" so the angle is not killed by mistake |
| E | Social proof | 21,740 reviews at 4.6 stars — let me settle the sizing question once. | 20.5% | Tests whether the "Social proof" angle itself holds in this category / market |
| F | Social proof | I have worn this for a month. Here is the honest version. | 12.3% | Same angle, different hook — separates "the angle failed" from "the hook failed" so the angle is not killed by mistake |
Kill criteria
- 3-second view-through < 25%: a hook problem. Test another hook on the same angle; do not touch the angle
- 3-second view-through fine but CTR < 0.8%: the middle is not persuading. Go back to the proof beat and add concrete evidence
- CTR fine but CVR < 1%: the problem is the landing page, the listing or the price — not the creative. Do not make new creative yet
- A cell spends its $30 with no orders: stop it and move the budget to the other hook on the same angle
Scale criteria
- ROAS > 1.2× target with $30 spent: add 50% budget to that cell and immediately produce 2 new hooks on the same angle as a relay
- Both hooks on one angle win: that is an "angle win" signal — extend the angle to the other language and another platform
- One hook wins but collapses on another angle: the hook won, not the angle — port that hook to the other angles
5. ROAS review sheet (fill in after the flight)
| Cell | Angle | Platform | Language | Spend | Impressions | 3s view-through | CTR | Orders | ROAS | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|
| A | Myth-bust | |||||||||
| B | Myth-bust | |||||||||
| C | Unboxing / Review | |||||||||
| D | Unboxing / Review | |||||||||
| E | Social proof | |||||||||
| F | Social proof |
Three questions for the review
- Did the angle win or did the hook? (Both hooks up = the angle won. Only one = the hook won.)
- Did the losing angle fail the category or the market? (Check the same angle in the other language.)
- What does this round change in the category knowledge base? (Which prior moves, which hook joins the library.)
6. Build info and render-cost estimate
- Engine
fixture+inference@v0.1 (no live fetching) + rule-engine@v0.1 (category knowledge base; no model called)
| Vendor | Shots / seconds | Estimated cost |
|---|---|---|
| HeyGen Avatar V (talking head) | 57 | $2.84 |
| Kling 3.0 (macro / B-roll) | 9 | $7.20 |
| Veo/Sora (hero shots) | 1 | $6.00 |
| Total (3 angles × 2 languages) | $16.04 |
Run it on your own SKU
Scan your listing copy first (free, no login)See the other samples
FAQ
Was this plan generated by an AI model?
No. This version comes from a rule engine plus the category knowledge base, with no model call — which is why the same link always produces the same plan, reviewable and diffable. When a model is wired in, the knowledge base still constrains the skeleton, the model only polishes and localises, and the risk flags still backstop it.
Why is the product fictional?
Because we do not want a real seller's product tied to advertising advice we invented. Category, price band, spec ranges and review counts follow real ranges; only the title and shop name are made up. Use your own link for your own SKU.
Did you plant those risky selling points on purpose?
Yes. Those lines are what real listings actually say. That is exactly where the risk flags earn their keep: they scan not our generated script but the seller's own sentences, the ones about to be copied into a voiceover and a creator video.
Can I shoot straight from this brief?
Follow the structure, not the content — the product in it does not exist. Take the template and fill in your own SKU, or run a plan on your own link.