7532536003957951758 - Social Video AI 分析报告

READY TO REUSE

完整口播脚本

Hello. We're gonna test out this wig grip and see how good it is by doing a shake test. Yeah, let's go. So obviously the bangs got a little bit messed up, but they're fine now, and that's not the fault of the wig grip. Obviously. Wig grips are not for bangs. Wigs. Wig grips over making sure the wig stays in place. My bangs are kind of precariously, um, set in place, so that's why they're. They're fixed now. They're okay. As you saw. Very effective wig grip. Wig did not move. I also can hardly take off this wig. I had to try to, like, I put it on the first time and I went, oh, I need to take it off so I can readjust it a little bit. And I hardly could because it's stuck to the wig grip. So good that I had to take the wig grip off with the wig, and then put the wig grip back on, and then put the wig back on. Very good wig grip. I need one. This one keeps falling off my head. Recommended by Big Chew. Friend should send me one, too. Haha, I'm just kidding. I'll buy one. 10 out of 10. You know the demon Mark flickering because of the filter? I've got a really bad pimple that I don't want you to see.

Hello. We're gonna test out this wig grip and see how good it is
by doing a shake test. Yeah, let's go.
So obviously the bangs got a little bit messed up,
but they're fine now, and that's not the fault of the wig grip.
Obviously. Wig grips are not for bangs.
Wigs. Wig grips over making sure the wig stays in place.
My bangs are kind of precariously, um,
set in place, so that's why they're.
They're fixed now. They're okay.
As you saw. Very effective wig grip.
Wig did not move. I also can hardly take off this wig.
I had to try to, like,
I put it on the first time and I went, oh,
I need to take it off so I can readjust it a little bit.
And I hardly could because it's stuck to the wig grip.
So good that I had to take the wig grip off with the wig,
and then put the wig grip back on,
and then put the wig back on.
Very good wig grip. I need one.
This one keeps falling off my head.
Recommended by Big Chew. Friend should send me one, too.
Haha,
I'm just kidding. I'll buy one.
10 out of 10. You know
the demon Mark flickering because of the filter?
I've got a really bad pimple that I don't want you to see.

二次判断

带入 Agent 继续分析

VERDICT

这条广告素材当前表现偏强，当前已经形成“结果展示”型表达；最值得保留的是“前 3 秒结果展示直接”，最需要优先修的是“CTA 信号偏弱”；当前 CTA 还偏弱，测试方式有说服力。

EVIDENCE

广告侧：2s 25.87% / 6s 8.67% / CTR 0.96% / ROI 21.59 / 41 单；00:01 就出现产品画面，动作是“presenter adjusts wig and shows hairline”，用户很快能知道卖点指向。

RISK

前 3 秒关键信息主要靠口播，没有同步强字幕，静音刷到时的停留效率会打折。

NEXT MOVE

测试方式已经具备说服力，下一步更适合围绕同类测试动作做不同开头和不同价格表达的变体。

规则判断用于快速复盘；进入 Agent 后可调用外部模型继续结合画面、脚本与市场语境追问。

Hook 复盘

结果展示型 Hook

前 3 秒核心判断

当前判为「结果先行」。判断只参考开头 3-5 秒的时间轴转写和首屏 OCR，避免后段证明动作误判成 Hook。；开头任务是先展示结果、强利益点或稳定效果。；开头口播/字幕依据：Hello. We're gonna test out this wig grip and see how good it is；产品在开头很早出现，适合直接建立广告理解。；后段可承接的证明动作：摇头/甩头稳定性测试、实测验证、close-up fit demonstration

这条素材做得好的点

加宽带在 1.0 秒内就明确露出，前几秒就完成品类识别，适合 TikTok 快节奏带货素材。
视频里有明确的稳定性测试或实测动作，这类证明镜头通常比纯讲解更能帮助转化。
镜前演示自带生活感，比较贴合同类 UGC 观看语境。
真实广告数据已经证明这条素材具备放量潜力，不只是内容好看，而是有实际转化支撑。
前 3 秒结果展示直接

最需要优先修的点

前 3 秒关键信息主要靠口播，没有同步强字幕，静音刷到时的停留效率会打折。
模型内容分不低，但真实完播偏弱，说明节奏感可能还是偏长或前几秒停留抓力不足。
CTA 识别偏弱，用户看完后未必知道该做什么动作。
当前单量成本高于同商品基准，说明这条创意还没有打到更优的成交效率。
CTA 信号偏弱

时间轴拆解

按片段聚合展示

00:00-00:08 开头 Hook

进入测试 / 证明动作

Hello. We're gonna test out this wig grip and see how good it is by doing a shake test. Yeah, let's go.

00:13-00:21 问题铺垫

产品结构与使用背景

So obviously the bangs got a little bit messed up, but they're fine now, and that's not the fault of the wig grip. Obviously. Wig grips are not for bangs.

00:21-00:27 问题铺垫

产品结构与使用背景

Wigs. Wig grips over making sure the wig stays in place. My bangs are kind of precariously, um, set in place, so that's why they're.

00:27-00:34 证明展示

上身效果与稳定性展示

They're fixed now. They're okay. As you saw. Very effective wig grip. Wig did not move. I also can hardly take off this wig.

00:34-00:38 证明展示

上身效果与稳定性展示

I had to try to, like, I put it on the first time and I went, oh, I need to take it off so I can readjust it a little bit.

00:38-00:45 证明展示

上身效果与稳定性展示

And I hardly could because it's stuck to the wig grip. So good that I had to take the wig grip off with the wig, and then put the wig grip back on,

00:45-00:56 证明展示

上身效果与稳定性展示

and then put the wig back on. Very good wig grip. I need one. This one keeps falling off my head. Recommended by Big Chew. Friend should send me one, too.

00:56-01:04 结果强化

结果强化与补充说明

Haha, I'm just kidding. I'll buy one. 10 out of 10. You know the demon Mark flickering because of the filter?

01:04-01:06 结尾引导

收尾总结

I've got a really bad pimple that I don't want you to see.

关键帧画面拆解

真实截图优先

00:01 产品已入镜

bedroom mirror setup

Frame hook-frame.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration
画面风险：text overlay slightly dense

00:05 产品未明确露出

bedroom mirror setup

Frame scene-01.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

00:13 产品未明确露出

bedroom mirror setup

Frame scene-02.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

00:24 产品未明确露出

bedroom mirror setup

Frame scene-03.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

00:33 产品未明确露出

bedroom mirror setup

Frame middle-frame.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

00:36 产品未明确露出

bedroom mirror setup

Frame scene-04.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

00:49 产品未明确露出

bedroom mirror setup

Frame scene-05.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

01:00 产品未明确露出

bedroom mirror setup

Frame scene-06.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

01:05 产品未明确露出

bedroom mirror setup

Frame ending-frame.png shows early product reveal and natural speaking style.

人物：one female presenter
动作：presenter adjusts wig and shows hairline
证明动作：close-up fit demonstration

定向优化建议

结合画面、口播与市场逻辑

可直接交给剪辑的建议

把 shake test 或稳定性测试剪进前中段，同时叠加一句更短的强字幕。
前 3 秒补一句更短的强字幕，比如“Hello. We're gonna test out”，优先把痛点、结果或题材关键词变成一眼能扫懂的文案。
保留镜前真人演示，但再补一个更近的局部镜头，避免全程停留在中景讲解。
既然视频强调发际线效果，建议单独补 1 秒发际线近景，并用字幕点明自然度。
当前是结果切入，建议首帧就把最强结果放在最显眼的位置。
缩短前 5 秒解释时长，把最能证明结果的镜头提到更前面，减少铺垫型口播。
结尾补一个明确 CTA：例如“点购物车看颜色”或“链接在下方”，不要只停在展示效果。
优先保留有效证明段，重做前 3 秒和 CTA，目标是把成本压回该商品的历史正常区间。

复剪价值判断

是否值得复剪：yes

原投放状态：已移除

测试方式识别

主表标签 + 视频二次识别

最终识别结果

主表原始测试方式

视频二次推断

视觉资产分析

不是只看截图，而是判断这帧为什么有用

00:01 开头画面

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：开头就能让用户知道卖什么，首屏识别成本低，适合做前 3 秒抓手。

00:05 开头画面

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：开头就能让用户知道卖什么，首屏识别成本低，适合做前 3 秒抓手。

00:13 使用步骤

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：发际线近景有助于传递自然度，适合做信任建立镜头。

00:24 使用步骤

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：发际线近景有助于传递自然度，适合做信任建立镜头。

00:33 使用步骤

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：发际线近景有助于传递自然度，适合做信任建立镜头。

00:36 使用步骤

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：发际线近景有助于传递自然度，适合做信任建立镜头。

00:49 使用步骤

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：发际线近景有助于传递自然度，适合做信任建立镜头。

01:00 使用步骤

产品

场景：bedroom mirror setup
机位：平视
景别：大特写
产品位置：中心清晰露出
画面理解：发际线近景有助于传递自然度，适合做信任建立镜头。

CTA 分析

成交引导是否清楚、是否带优惠和紧迫感

CTA 文案 I'm just kidding. I'll buy one.

CTA 出现时间 00:57

CTA 类型 未识别

CTA 强度 weak

是否有优惠否

是否有紧迫感否

产品露出分析

不只是出现没出现，而是出现得够不够早、够不够清楚

产品首次出现 00:01

总露出时长 13.50 秒

产品近景次数 9

产品使用清晰度高

产品露出评分 100

Logo 可见否

包装可见否

价格可见否

Commercial Reality Check

内容表现之外，再看真实成交反馈

数据侧结论

加宽带已经出现早期成交信号。最近一期 GMV 约 16.99，订单 1，滚动三期 GMV 约 16.99。当前是新近出现的成交信号，值得继续观察是否能放大。当前退款件数占成交件数约 0.00%。

趋势判断

趋势：new_positive

滚动周期：1 个阶段

最近一期 0326-0425

最近一期 GMV $16.99

周期	GMV	订单数	成交件数	退款金额	退款件数
0326-0425	$16.99	1	1	$0.00

Ad Performance Layer

人工拉取广告数据后写入的真实表现层

数据侧判断

这条素材在广告投放侧已有真实信号：ROI 21.59，CTR 0.96%，转化率 14.89%，完播率 0.30%。说明它不仅读得动点击，也具备较强成交潜力，适合优先当作高价值模板拆解。加宽带在手动广告数据里也有商品级基准：累计营收 500,659.90，订单 29795，ROI 4.20。它的投放表现已经不低于当前商品级基准，可以优先沉淀成高价值样本。

数据来源说明

创意类型：视频
视频来源：-
投放状态：不可用
发布时间：2025-07-30 00:07
商品基准分类：wide_band_video_perf
商品基准名称：加宽带

视频广告 ROI 21.59

广告花费 $30.15

广告收入 $650.83

订单数 41

曝光 4895

点击 47

CTR 0.96%

转化率 14.89%

2s 留存 25.87%

6s 留存 8.67%

25% 播放率 3.31%

50% 播放率 1.27%

75% 播放率 0.81%

完播率 0.30%

商品基准 ROI 4.20

商品基准营收 $500,659.90

商品基准订单 29795

商品平均下单成本 $4.00

表现归因与下一步动作

把内容判断翻译成更可执行的动作建议

这条广告素材当前表现偏强，当前已经形成“结果展示”型表达；最值得保留的是“前 3 秒结果展示直接”，最需要优先修的是“CTA 信号偏弱”；当前 CTA 还偏弱，测试方式有说服力。

表现归因

当前判断：当前表现偏强，适合重点保留

前 3 秒结果展示直接
产品露出足够早
有明确证明动作
真人生活感较强
测试方式具体，便于用户快速建立信任
展示路径完整，用户更容易理解产品怎么用

图文证据

00:01 就出现产品画面，动作是“presenter adjusts wig and shows hairline”，用户很快能知道卖点指向。
中段已经出现摇头/甩头稳定性测试、实测验证，说明这条视频不是只讲卖点，而是有在给用户看证据。
前 3 秒口播重点是“Hello. We're gonna test out this wig grip and see how...”，这直接决定了用户第一反应更像在接收痛点、结果还是教程。

主要拖累点

CTA 信号偏弱

数据反馈

广告侧真实信号：ROI 21.59、CTR 0.96%、转化率 14.89%、完播率 0.30%。

下一步动作

建议动作：做轻变体测试

处理优先级：中，建议排在下一轮优化

具体做法：测试方式已经具备说服力，下一步更适合围绕同类测试动作做不同开头和不同价格表达的变体。

为什么先做这个

因为测试动作本身就有说服力，下一步最值得做的是保留测试中段，围绕不同开头、不同价格表达和不同 CTA 做轻变体。

历史相似参考样本

把当前视频和已沉淀素材库做对照，避免建议只停留在模板话术

这一栏不是让我们照抄历史素材，而是快速判断：当前视频像谁、能借鉴哪种证明动作或镜头节奏、哪些地方需要做差异化。

7512873337257872683

达人：mykengoin · 产品：加宽带

相似度 105 结果展示型 Hook

为什么相似

共同命中的展示/测试方式是：展示发际线、展示穿戴过程、甩头；Hook 结构同为 result_showcase，但这只作为辅助参考，不等同于同一种创意。；同属加宽带，用户理解成本接近。；产品首次露出节奏接近，相差约 0.0 秒。；历史样本开头信息是：you all are gobbling up these wig grips from Dream Lover and for great reason they are so

可以借鉴什么

开头用的是“you all are gobbling up these wig grips from Dream Lover and for great”，可以参考它先抛出具体场景/感受，而不是只报产品名；可借鉴的动作是“甩头、扯头、梳头”，重点看动作结果有没有被镜头拍清楚；有摇头或稳定性测试，证明动作比纯讲解更容易带来信任感。

差异化提醒

相似样本只提供参考方向，当前视频仍要根据自己的首帧、口播和产品露出位置做差异化。

画面参考点

开头就能让用户知道卖什么，首屏识别成本低，适合做前 3 秒抓手。

内容评分97

UGC 真实感78

建议动作补证明镜头

新窗口打开历史报告

7547863217343630623

达人：missgiad · 产品：加宽带

相似度 105 结果展示型 Hook

为什么相似

共同命中的展示/测试方式是：展示发际线、展示穿戴过程、甩头；Hook 结构同为 result_showcase，但这只作为辅助参考，不等同于同一种创意。；同属加宽带，用户理解成本接近。；产品首次露出节奏接近，相差约 0.0 秒。；历史样本开头信息是：alright so I've got the Dream Lover Wig Grip in today

可以借鉴什么

开头用的是“alright so I've got the Dream Lover Wig Grip in today”，可以参考它先抛出具体场景/感受，而不是只报产品名；可借鉴的动作是“甩头、扯头、梳头”，重点看动作结果有没有被镜头拍清楚；有摇头或稳定性测试，证明动作比纯讲解更容易带来信任感。

差异化提醒

相似样本只提供参考方向，当前视频仍要根据自己的首帧、口播和产品露出位置做差异化。

画面参考点

开头就能让用户知道卖什么，首屏识别成本低，适合做前 3 秒抓手。

内容评分94

UGC 真实感84

建议动作继续放量

新窗口打开历史报告

7556674200602201375

达人：unbrandied · 产品：加宽带

相似度 100 结果展示型 Hook

为什么相似

共同命中的展示/测试方式是：展示发际线、展示穿戴过程；Hook 结构同为 result_showcase，但这只作为辅助参考，不等同于同一种创意。；同属加宽带，用户理解成本接近。；产品首次露出节奏接近，相差约 0.0 秒。；历史样本开头信息是：Look at my new wig grip that. Found this one on the TikTok Shop. This is the one that is t

可以借鉴什么

开头用的是“Look at my new wig grip that. Found this one on the TikTok Shop. This ”，可以参考它先抛出具体场景/感受，而不是只报产品名；可借鉴的动作是“展示穿戴过程、展示发际线、颜色选择”，重点看动作结果有没有被镜头拍清楚；已经出现证明动作：佩戴舒适度证明、结构细节展示、close-up fit demonstration。

差异化提醒

相似样本只提供参考方向，当前视频仍要根据自己的首帧、口播和产品露出位置做差异化。

画面参考点

开头就能让用户知道卖什么，首屏识别成本低，适合做前 3 秒抓手。

内容评分100

UGC 真实感86

建议动作补证明镜头

新窗口打开历史报告

打法规则库引用

把单条素材放回 172 条样本沉淀出的规则里校准

样本数 499

高质量样本 303

强证据样本 374

3秒内产品露出率 100.0%

hook_first_3s

前 3 秒要同时解决看懂和想看

规则证据

样本中 455/455 条可判定产品露出的素材在 3 秒内露出产品。
高频拖累项包含：前 3 秒抓手还不够集中, 证明动作不够强, CTA 信号偏弱。

执行建议

不要只靠口播解释，首屏要用字幕或画面直接表达痛点、结果或测试动作。
如果产品早露出但留存弱，优先补强第一句话和首帧封面，而不是只加结尾 CTA。

proof_action

防滑类素材必须有可视化证明动作

规则证据

高频测试方式包括：展示发际线, 展示穿戴过程, 扯头, 颜色选择, 梳头, 甩头。
高频优势项包括：真人生活感较强, 展示路径完整，用户更容易理解产品怎么用, 产品露出足够早, 测试方式具体，便于用户快速建立信任。

执行建议

甩头、扯头、梳头、展示发际线、展示穿戴过程要尽量出现在中前段。
证明动作不要只说，要让用户看到动作结果，比如拉扯后不移位、梳头后发际线自然。

cta_close_loop

内容看懂之后要有明确行动收口

规则证据

CTA 强度达到 medium/strong 的样本占比约 46.1%。
当前 next_action 高频项：add_proof, scale, keep_testing, unknown, make_variations。

执行建议

如果视频有证明动作但成交弱，优先检查 CTA 是否讲清楚在哪里买、为什么现在买。
折扣、TikTok Shop、评论区引导和链接下方提示最好结合场景自然出现，不要像硬广口号。

visual_asset_reasoning

关键帧不是截图库存，而是画面判断力

规则证据

高频画面类型：使用步骤, 开头画面, 过程画面。
高频主体：产品。

执行建议

优先挑一眼能看懂产品用途、痛点或证明动作的帧做封面候选。
如果画面主体是人但产品不清楚，报告建议应偏向补产品近景；如果产品清楚但人不真实，建议偏向补真人生活感。

打法库给这条的额外提醒

打法库提示：这条素材可以优先考虑补一个更可视化的证明动作，比如扯头 / 颜色选择，但要结合当前脚本自然插入，不能硬塞测试镜头。

过泛标签提醒

前 3 秒抓手还不够集中：91.2%。出现比例过高，暂时不适合作为区分好坏素材的核心判断，需要后续拆成更细的原因。
证明动作不够强：88.4%。出现比例过高，暂时不适合作为区分好坏素材的核心判断，需要后续拆成更细的原因。
产品露出足够早：87.8%。出现比例过高，更像基础特征，不宜直接当成强优势，需要和数据表现或画面证据一起判断。
真人生活感较强：91.2%。出现比例过高，更像基础特征，不宜直接当成强优势，需要和数据表现或画面证据一起判断。
展示路径完整，用户更容易理解产品怎么用：90.6%。出现比例过高，更像基础特征，不宜直接当成强优势，需要和数据表现或画面证据一起判断。

转写、OCR 与音乐补充

默认折叠，方便复核

完整口播转写

Hello. We're gonna test out this wig grip and see how good it is by doing a shake test. Yeah, let's go. So obviously the bangs got a little bit messed up, but they're fine now, and that's not the fault of the wig grip. Obviously. Wig grips are not for bangs. Wigs. Wig grips over making sure the wig stays in place. My bangs are kind of precariously, um, set in place, so that's why they're. They're fixed now. They're okay. As you saw. Very effective wig grip. Wig did not move. I also can hardly take off this wig. I had to try to, like, I put it on the first time and I went, oh, I need to take it off so I can readjust it a little bit. And I hardly could because it's stuck to the wig grip. So good that I had to take the wig grip off with the wig, and then put the wig grip back on, and then put the wig back on. Very good wig grip. I need one. This one keeps falling off my head. Recommended by Big Chew. Friend should send me one, too. Haha, I'm just kidding. I'll buy one. 10 out of 10. You know the demon Mark flickering because of the filter? I've got a really bad pimple that I don't want you to see.

脚本稿整理版

Hello. We're gonna test out this wig grip and see how good it is
by doing a shake test. Yeah, let's go.
So obviously the bangs got a little bit messed up,
but they're fine now, and that's not the fault of the wig grip.
Obviously. Wig grips are not for bangs.
Wigs. Wig grips over making sure the wig stays in place.
My bangs are kind of precariously, um,
set in place, so that's why they're.
They're fixed now. They're okay.
As you saw. Very effective wig grip.
Wig did not move. I also can hardly take off this wig.
I had to try to, like,
I put it on the first time and I went, oh,
I need to take it off so I can readjust it a little bit.
And I hardly could because it's stuck to the wig grip.
So good that I had to take the wig grip off with the wig,
and then put the wig grip back on,
and then put the wig back on.
Very good wig grip. I need one.
This one keeps falling off my head.
Recommended by Big Chew. Friend should send me one, too.
Haha,
I'm just kidding. I'll buy one.
10 out of 10. You know
the demon Mark flickering because of the filter?
I've got a really bad pimple that I don't want you to see.

分段时间轴转写

00:00-00:05
Hello. We're gonna test out this wig grip and see how good it is
00:05-00:08
by doing a shake test. Yeah, let's go.
00:13-00:16
So obviously the bangs got a little bit messed up,
00:16-00:18
but they're fine now, and that's not the fault of the wig grip.
00:18-00:21
Obviously. Wig grips are not for bangs.
00:21-00:23
Wigs. Wig grips over making sure the wig stays in place.
00:23-00:26
My bangs are kind of precariously, um,
00:26-00:27
set in place, so that's why they're.
00:27-00:29
They're fixed now. They're okay.
00:29-00:31
As you saw. Very effective wig grip.
00:31-00:34
Wig did not move. I also can hardly take off this wig.
00:34-00:34
I had to try to, like,
00:34-00:36
I put it on the first time and I went, oh,
00:36-00:38
I need to take it off so I can readjust it a little bit.
00:38-00:41
And I hardly could because it's stuck to the wig grip.
00:41-00:43
So good that I had to take the wig grip off with the wig,
00:43-00:45
and then put the wig grip back on,
00:45-00:46
and then put the wig back on.
00:46-00:48
Very good wig grip. I need one.
00:48-00:50
This one keeps falling off my head.
00:52-00:56
Recommended by Big Chew. Friend should send me one, too.
00:56-00:57
Haha,
00:57-00:59
I'm just kidding. I'll buy one.
01:00-01:02
10 out of 10. You know
01:02-01:04
the demon Mark flickering because of the filter?
01:04-01:06
I've got a really bad pimple that I don't want you to see.

OCR 字幕汇总

暂无 OCR 字幕。

音乐分析

暂无音乐分析结果。

短期与长期学习系统预留

后续会根据这条素材的分析时间 2026-07-04 08:46:09 UTC 关联最近热点与高转化素材，补一层“本周市场提醒”。
后续会把已分析素材沉淀进长期知识库，用于新素材与历史优秀素材的相似度对比、复剪参考与团队教学。

Turn Raw Social Video Into Usable Creative Insight

完整口播脚本

二次判断

这条广告素材当前表现偏强，当前已经形成“结果展示”型表达；最值得保留的是“前 3 秒结果展示直接”，最需要优先修的是“CTA 信号偏弱”；当前 CTA 还偏弱，测试方式有说服力。

Hook 复盘

前 3 秒核心判断

这条素材做得好的点

最需要优先修的点

时间轴拆解

进入测试 / 证明动作

产品结构与使用背景

产品结构与使用背景

上身效果与稳定性展示

上身效果与稳定性展示

上身效果与稳定性展示

上身效果与稳定性展示

结果强化与补充说明

收尾总结

关键帧画面拆解

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

bedroom mirror setup

定向优化建议

可直接交给剪辑的建议

复剪价值判断

测试方式识别

最终识别结果

主表原始测试方式

视频二次推断

视觉资产分析

产品

产品

产品

产品

产品

产品

产品

产品

CTA 分析

产品露出分析

Commercial Reality Check

数据侧结论

趋势判断

Ad Performance Layer

数据侧判断

数据来源说明

表现归因与下一步动作

表现归因

图文证据

主要拖累点

数据反馈

下一步动作

为什么先做这个

历史相似参考样本

为什么相似

可以借鉴什么

差异化提醒

画面参考点

为什么相似

可以借鉴什么

差异化提醒

画面参考点

为什么相似

可以借鉴什么

差异化提醒

画面参考点

打法规则库引用

前 3 秒要同时解决看懂和想看

规则证据

执行建议

防滑类素材必须有可视化证明动作

规则证据

执行建议

内容看懂之后要有明确行动收口