📡 Idea Radar 每天 8:00 / 18:00 推送 · 18:00 生成精选
reddit · SideProject · 33 分 ← 返回列表

YouTube is too long, TikTok is too shallow. So I built the middle.

We all live the same thing, right? > > There's an amazing 2-hour documentary / podcast on YouTube, but you don't have time to watch it. You open TikTok instead, you scroll for 2 hours, and you remember literally nothing. > > So I built the middle: > > I take long, high-quality YouTube videos and condense them into their most impactful moments. Most videos are just 2-5 minutes, and the longest ones are 15 minutes max. And instead of vertical scrolling, you swipe horizontally -

五维评分

付费意愿
20
需求真实度
30
个人可落地
55
竞争空间
25
新鲜度
50

AI 分析

痛点:长视频(播客、纪录片)内容质量高但没时间看,短视频刷完又什么都记不住,中间缺少一个能把 2 小时内容概括成 2-15 分钟的优质内容发现与消费形态。

目标用户:有自我提升需求的年轻知识工作者、学生、通勤族;内容偏好播客、纪录片、历史类长视频。

证据:There's an amazing 2-hour documentary / podcast on YouTube, but you don't have time to watch it. You open TikTok instead, you scroll for 2 hours, and you remember literally nothing. Do you think this could work? What kind of videos would you want summarized most?

变现:订阅制(免费看若干条/天,付费解锁无限与离线);按领域付费专题包;与播客/创作者做联盟分成或导流分成。

MVP:先做一个 Web 或 MVP App,爬取指定 YouTube 频道新视频,用现成 LLM 或 ASR 管线生成 2-5 分钟高光切片并人工挑选把关,横向滑动卡片式信息流;只做 3-5 个垂直领域,用邮件或邀请码限量发放,验证每个用户每周打开几屏、第二天是否回访。

竞品:Eightify、Opus Clip、Snipd、YouTube Summarized、Podwise、Glasp

摘要:作者发布了把长 YouTube 视频压缩成 2-15 分钟高光片段、并以横向滑动信息流呈现的 Side Project,理由是从 2 小时播客到 TikTok 之间缺少一个中间形态。帖子只是在征求反馈(“Do you think this could work?”),没有任何付费意愿表述、收入数字或用户附和的证据,本质上仍只是创作者自己对需求的假设;同时 AI 摘要、切片赛道已有大量竞品,销售空间有限,因此只适合作为低成本验证型的小项目。