素材验收记分表
更新于 2026-08-15 · 风险词表版本 2026-08-15c
达人回片之后、花钱之前用的一张表。做法沿用一条用户研究的纪律:只有可观察的东西算分,"挺好看的"记 0。先过 7 条硬性检查,再打 0–10 分决定投不投。
全文在下面,也可以下载 .md。免费、可商用、可改、不用署名。
模板是英文的:收件人(达人 / 拍摄方)多半读英文,中文卖家看结构就够了。
下载 creative-scorecard.md纯文本 Markdown,可直接贴进飞书 / Notion / 企微,也可以打印。
这份模板里有什么
| 节 | 内容 |
|---|---|
| A. Brief compliance (pass/fail — any fail sends it back) | |
| B. Score (0–10) | |
| C. What we do next | |
| D. Notes back to the creator |
怎么用
- A 段是一票否决,不是打分。任何一条没打勾就退回重剪或重拍——不要"在字幕里补一句"糊过去,那句字幕本身就是风险。
- B 段刻意有负分。最强的一下出现在第 15 秒之后要扣分,因为没有人看到那里;这一条比任何"建议前置高潮"的说法都管用。
- "能不能只重拍一个镜头"是这张表最省钱的用法。C 段刻意不给"全部重来"这个选项。
- 纪律来源:这套"只看可观察行为、态度一律记 0"的打分法是我们做卖家访谈时用的判定方式,搬到素材验收上同样成立——两处失效的方式是同一种:把客气话当成信号。
模板全文
下面就是完整内容(英文,因为收件人多半读英文)。可以直接全选复制,也可以用上面的下载按钮。
# Creative acceptance scorecard > Score the video you got back **before** you spend money on it. > Rule borrowed from user research: only observable things score. "It looks nice" is zero. Video: [file / link] · Creator: [name] · SKU: [name] · Date: [date] · Scored by: [name] ## A. Brief compliance (pass/fail — any fail sends it back) | # | Check | Pass? | Note | |---|---|---|---| | A1 | The hook is said verbatim in the first 3 seconds | ☐ | | | A2 | Nothing from the "do not say" list appears in speech | ☐ | | | A3 | Nothing from the "do not say" list appears in captions | ☐ | | | A4 | No scales, tape measures, body measurements, or pinching fat | ☐ | | | A5 | Before/after (if used) changed exactly one variable | ☐ | | | A6 | Aspect and safe area are correct | ☐ | | | A7 | AI-generated content is disclosed, if applicable | ☐ | | **Any unchecked box above = send back for a re-shoot or a re-edit. Do not "fix it in the caption".** ## B. Score (0–10) | Points | Observable thing | |---|---| | +3 | The proof shot exists, is one take, and is legible on a phone at arm's length | | +2 | The first 3 seconds contain a physical action, not a talking head | | +2 | The claim made on screen is something the garment actually does | | +1 | The size worn is stated, with the wearer's height/weight or usual size | | +1 | Audio is usable without music (you can hear the words) | | +1 | The video still makes sense with the sound off | | −2 | The strongest moment is after second 15 (nobody gets there) | | −2 | The product is only visible flat-lay / on a hanger, never worn in motion | | −1 | The creator reads the script instead of talking | | −1 | The caption repeats the voiceover word for word | **Reading**: 7+ run it · 4–6 run it in the cheapest test cell only · ≤3 do not spend on it. ## C. What we do next - [ ] Runs as A/B cell: [which cell / which hook it is testing] - [ ] Re-cut needed: [what exactly] - [ ] Re-shoot needed: [which shot only — never "do it all again"] - [ ] Pay / repost / send the next sample ## D. Notes back to the creator Say what to keep, not just what to fix. One line each: - Keep: [what worked] - Change: [one thing, the most important one] --- 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. Risk-term list version 2026-08-15c. Template published by Preshoot — free to copy, edit and use commercially.
其它模板
达人 brief 模板(英文,一页纸)一页纸的英文达人 brief。填完括号就能直接发出去,含"不要说的话"和"不要拍的画面"两节——寄样之前把这两节写清楚,比拍完再返工便宜得多。ROAS 复盘表测试窗口结束后填的表。它只回答一个问题:赢的是角度还是钩子——这决定下一笔预算该换角度还是换第一句话。停投与放量的阈值要求你在投之前就填好。
常见问题
为什么要在花钱之前打分?
因为投出去之后再判断,你分不清是素材差还是投放差。先用一张表把明显不能投的挡掉,剩下的差异才有归因价值。
7 分以上就一定能跑赢吗?
不能。这张表挡的是"明显浪费钱的素材",不预测赢面。赢面要靠 A/B 矩阵和复盘表测出来。
达人会不会觉得被挑刺?
把 D 段("保留什么 / 改哪一件事")真的填了就不会。只说改什么的反馈会让人防御,先说保留什么的反馈会让人改。
为什么"字幕和口播逐字相同"要扣分?
因为屏幕上的字和耳朵里的话重复,等于浪费了一条信息通道。字幕该写口播没说的那一半。