我们因为一个没有收入的副项目被罚款6万美元。
在2020年，正值新冠疫情期间，我和几个朋友决定，显然世界需要一种<i>新的</i>演示工具。
我们的痛点很简单：我们讨厌与PowerPoint/Google Slides斗争。我们只想写内容，得到漂亮的幻灯片，而不必操心字体、大小、对齐、颜色，以及19种看起来都很糟糕的“标题+副标题”布局。
于是我们构建了一个原型：一个小型的点击式幻灯片生成器。然后我们觉得点击太慢，于是将其转变为一个完整的Markdown转幻灯片的工具。输入Markdown，输出漂亮的幻灯片。简单。
接着，像所有好的副项目一样，我们开始添加功能：图片上传、通过API生成PDF、AI集成，以及越来越多没人真正要求的功能……
在这个过程中，在“我们绝对是下一个独角兽”的巅峰状态下，我们通过Stripe Atlas注册了一家美国公司。为了显得更正式，我们给自己发行了10,000,000股。因为这就是所谓的真正创业公司该做的，对吧？
然后我们又回去高高兴兴地发布新功能……而且不那么高兴地没有推出任何付费计划。
（快进大约一年半）
有一天，一个同事给我们发消息：
> “嘿，联邦门户网站说我们因未提交税务申报表而欠60,000美元的罚款。”
我们都感受到了一种通常只在生产数据库崩溃时才会出现的全身热浪。
（给读者的说明：我们最初来自欧盟，在那里如果公司没有收入，就不需要缴纳任何税款，如果收到罚款，通常也不过几百欧元而已）
于是我们学到了一些有趣的事情：你发行的股份数量会影响各种费用和罚款的计算方式。在你兴奋地点击10,000,000之前，知道这一点真是太好了。
我们花了几天几夜阅读IRS文档，给人发邮件，试图理解我们是否刚刚让一个甚至没有定价页面的项目破产。
最后，解决方案出乎意料地无聊：
1) 我们提交了一份零收入的税务申报表。
2) 在线系统重新计算了一切。
3) 我们可怕的60,000美元罚款变成了大约1,500美元，我们支付了。
危机算是暂时避免了。
然后下一年的年末临近，我们都想到同样的事情：“我们不想再这样做了。”
从技术上讲，我们可能可以选择忽视这家公司，让它慢慢走向法律的死亡。但我们是那些喜欢“妥善处理事情”的人，所以我们又花了大约2,000美元请人帮我们以正确的方式关闭公司。
我们“我们绝对是创始人了”的时刻总账单：大约3,500美元，伴随着大量的压力和0美元的收入。
之后，我们基本上让这个项目休眠。每个人都有自己的业务，而这次经历已经足够痛苦。
除了……我们从未停止使用它。
在内部，我们继续用它生成所有幻灯片。老实说？输出效果依然很好。速度快，外观好，一旦你习惯了，回到Google Slides或Keynote就像是在Excel中设计海报。
最近，AI帮助我们重构了一些代码，打磨了一些细节，并添加了我们一直想要但从未有精力去实现的功能。
所以现在这个项目在slidepicker.com上重新焕发生机。
从这一切中有两个要点：
1. 在你至少有一个付费客户之前，不要注册公司（尤其是在其他国家）。等这些客户产生稳定收入后再注册可能更好。
2. 如果你决定注册公司，先找一个真正懂税务的人咨询，而不是兴奋地在表格中输入“10,000,000股”。
如果你只想从Markdown中获得漂亮的幻灯片，而不想意外地快速了解美国公司法，可以在slidepicker.com上试试。
查看原文
In 2020, in the middle of COVID, a few friends and I decided that obviously the world needed <i>another</i> presentation tool.<p>Our pain point was simple: we hated fighting with PowerPoint/Google Slides. We just wanted to write content and get nice slides without babysitting fonts, sizes, alignment, colors, and 19 different "Title + Subtitle" layouts that all somehow looked bad.<p>So we built a prototype: a tiny click-together slide builder. Then we decided clicking was too slow and turned it into a full markdown-to-slides thing. Write markdown in, pretty slides out. Easy.<p>Then, like every good side project, we started bolting on features: Image uploads, PDF generation via API, AI integration and more and more stuff nobody had actually asked for....<p>Somewhere in there, in peak "we're definitely the next unicorn" mode, we incorporated a US company through Stripe Atlas. To feel extra serious, we issued ourselves 10,000,000 shares. Because that's what real startups do, right?<p>Then we went back to happily shipping features and... not so happily not. shipping a single paid plan.<p>(fast-forward about a year and a half)<p>One day a colleague messages us:<p>> "Hey, the federal portal says we owe $60,000 in penalties for not filing a tax return."<p>We all experienced the same full-body heat wave usually reserved for production database drops.<p>(Note for readers: We are originally from the EU, where if a company has no income, it doesn't pay any taxes and if you receive a fine, it is usually a few hundred euros max)<p>And so we learned something interesting: the number of shares you issue can influence how various fees and penalties are calculated. Which is nice to know BEFORE you click 10,000,000.<p>We spent days and nights reading IRS docs, emailing people, trying to understand if we'd just bankrupted a project that didn't even have a pricing page.<p>In the end, the solution was hilariously boring:<p>1) We filed a zero-revenue tax return.<p>2) The online system recalculated everything.<p>3) Our terrifying $60k penalty turned into something around $1500, which we paid.<p>Crisis kind of averted.<p>Then the end of the next year approached and we all had the same thought: "We're not doing this again."<p>Technically, we probably could have just ignored the company and let it die a slow legal death. But we're those annoying people who like things "done properly", so we paid roughly another $2,000 to have someone help us shut the company down the right way.<p>Total bill for our "we're definitely founders now" moment: aprox. $3,500 and a lot of stress and $0 in revenue.<p>After that, we basically put the project to sleep. Everyone had their own businesses, and this was a sufficiently painful lesson.<p>Except... we never stopped using it.<p>Internally, we kept generating all our slides with it. And honestly? The output still feels great. It's fast, it looks good, and once you get used to it, going back to Google Slides or Keynote feels like trying to design posters in Excel.<p>Recently, AI helped us refactor a bunch of the code, polish things, and add features we always wanted but never had the energy to implement.<p>So now the project is alive again at slidepicker.com.<p>The two takeaways from all this:<p>1. Don't incorporate (especially in another country), until you have at least one paying customer. It may be better to wait until these customers generate a stable income.<p>2. If you do incorporate, talk to someone who actually understands taxes before you excitedly type "10,000,000 shares" into a form.<p>And if you just want nice slides from markdown without accidentally speedrunning US corporate law, you can play with it at slidepicker.com.