reddit · SideProject · 10-19 12:50 47 分 ← 返回列表

I had 31 tabs open trying to buy an office chair, so I built something that only ever gives you three

I needed a desk chair. Three evenings later I had 31 tabs, a spreadsheet comparing warranties, and no chair. The problem wasn't a lack of information. It was that nothing would just tell me what to buy. So I built TierShop . You type in what you're buying and it gives you three results: the cheapest one actually worth owning, the one most people should buy, and the nice one. Compared on whatever actually differs for that category — chairs on adjustment range and warranty, baby wipes on cos

五维评分

付费意愿

需求真实度

个人可落地

竞争空间

新鲜度

AI 分析

痛点：买稍微复杂一点的商品时，信息过载会导致决策瘫痪：开了几十个标签页、做了对比表格，最后还是不知道买哪一个，没人愿意直接给出结论。

目标用户：需要快速做出购买决策的普通网购消费者，尤其是买办公椅、婴儿湿巾这类需要比较参数和保修的品类。

证据：I had 31 tabs open trying to buy an office chair... The problem wasn't a lack of information. It was that nothing would just tell me what to buy. / Honest state of it: five beta testers, one paying customer (me)... Every search reads the live web, which costs me real money per search, so there's no unlimited free tier

变现：采用按次搜索付费或订阅制（覆盖实时联网搜索成本），并叠加电商联盟佣金；高成本品类可做高价一次性深度报告。

MVP：两周内锁定1-2个参数复杂的品类（办公椅、婴儿湿巾），用搜索API+LLM做“三档推荐+明确赢家+拒绝不适用的品类”，用Stripe做按次付费或低价月订阅，先验证非本人付费转化。

竞品：Wirecutter、Perplexity、Amazon Rufus、ChatGPT购物推荐、各类AI选品/比价插件

摘要：作者为买办公椅开了31个标签页、做了对比表格仍未下单，于是做了TierShop：输入商品只给三档推荐并强制选出一个赢家，还会拒绝不适用品类。目前只有5个beta测试者、1个付费用户（作者本人），且每次搜索都要实时联网、有真实成本，因此付费需求尚未被外部验证。