聪明的创始人选择糟糕创业想法的原因
大多数创业建议乍一看似乎都很合理。“找到一个问题。” “解决自己的痛点。” “构建人们想要的东西。”
我遵循了这些建议，但仍然看到聪明的创始人，包括我自己，花了几个月时间构建的东西悄无声息地失败了。
这种矛盾让我感到困扰，因此我开始深入研究。我阅读了失败的事后分析、无果的产品发布、被遗弃的GitHub仓库，以及长长的Hacker News讨论串，里面的人们解释了为什么放弃某个想法。
经过一段时间，我发现了一个模式。问题不在于人们缺乏想法，而在于我们用讲故事的逻辑来评估想法，而不是用生存的逻辑。
我们选择那些听起来有趣、感觉雄心勃勃或在推介中看起来不错的想法，但一旦遇到现实世界的行为就会崩溃。
大多数想法失败的原因不是竞争或执行，而是它们并没有替代任何紧迫的需求。
它们无法与现有的习惯、重复的成本或某人已经在使用的痛苦的变通方法相匹配。
当你问一些简单的问题，比如谁在付费、他们在采用这个想法时停止做什么，以及为什么是现在，大多数想法很快就会崩溃。
为了避免重复这个错误，我开始将想法写成假设，而不是灵感。
每个想法都必须经受几个不舒服的问题：这个想法替代了什么现有行为、在前30天内会导致它失败的因素是什么，以及什么是可以证明或反驳需求的最小实验。
大多数想法立即失败。少数想法的存活时间超过了预期。
随着时间的推移，这变成了一个私人数据库，我用它来避免在弱想法上浪费数月时间。这不是一个“伟大想法”的集合。
它是经过严格筛选后存活下来的想法的记录，还有许多没有存活的想法。最终，我将其整理成一个其他人可以浏览的东西，现在称为startupideasdb.com（可以搜索一下），主要是因为我一直希望早些时候就有这样的东西。
如果你在选择构建什么时感到困惑，或者厌倦了那些悄然死去的聪明想法，这可能会为你节省一些时间。
如果你不同意这个框架，我真心希望了解它的缺陷，Hacker News通常能比其他地方更快发现问题。
查看原文
Most startup advice sounds reasonable on the surface. “Find a problem.” “Scratch your own itch.” “Build something people want.”
I followed all of it, and still watched smart founders, including myself, spend months building things that quietly went nowhere.<p>That contradiction bothered me enough to dig deeper. I started reading failure postmortems, dead Product Hunt launches, abandoned GitHub repos, and long Hacker News threads where people explain why they passed on an idea.<p>After a while, a pattern emerged. The problem isn’t that people lack ideas. It’s that we evaluate ideas using storytelling logic instead of survival logic.<p>We choose ideas that sound interesting, feel ambitious, or look good in a pitch, but collapse the moment they meet real-world behaviour.<p>What kills most ideas isn’t competition or execution. It’s that they don’t replace anything urgent.<p>They don’t map to an existing habit, a recurring cost, or a painful workaround someone is already using.<p>When you ask simple questions like who is paying, what they stop doing when they adopt this, and why now, most ideas fall apart very quickly.<p>To stop repeating this mistake, I began writing ideas down as hypotheses instead of inspiration.<p>Each idea had to survive a few uncomfortable questions: what existing behavior does this replace, what would kill it in the first thirty days, and what is the smallest experiment that could prove or disprove demand.<p>Most ideas failed immediately. A few survived longer than expected.<p>Over time this turned into a private database I used to avoid wasting months on weak ideas. It wasn’t a collection of “great ideas.”<p>It was a record of ideas that survived brutal filtering, along with many that didn’t. Eventually I cleaned it up into something others could browse, now called startupideasdbcom (google it), mostly because I kept wishing something like this existed earlier.<p>If you’re stuck choosing what to build, or tired of clever ideas that die quietly, this might save you some time.<p>And if you disagree with the framework, I’m genuinely interested in where it breaks, Hacker News usually finds the flaws faster than anywhere else.