做产品 PMaker
空格的键盘
所有需要管理项目的人 自由职业者 同时接多个客户的设计师 用微信跟客户对接、 每月要对账的设计师 说了等于没说 能直接推出功能

每收窄一层,你就多知道一件他一定会遇到的事。收到最后那层,界面基本自己就长出来了。

Narrow the target audience

The broader the user description, the fewer decisions it lets you make. Only when you can picture one specific person does the product take shape.

What you will run into
  • 做功能时总在纠结「有的人可能要,有的人可能不要」
  • 产品什么人都能用,但什么人都不会特意选它
  • 写文案时写不出具体的话,只能写「更高效」这类词

Broad user descriptions feel reassuring because they write the market size large. That number is an illusion — you cannot satisfy any two of those groups at once, because their situations are different.

The real benefit of narrowing is that it lets you subtract. If you know your users are on mobile, outdoors, wearing gloves, you can confidently drop a lot of design options, and what is left gets simpler on its own.

Which direction to narrow

Narrowing is not a random cut. Cut along the dimensions that would change how you build the product.

沿这个维度收它会改变什么
使用场合在工位上还是在路上,决定了屏幕大小、单手操作、能不能被打断
熟练程度老手要快捷键和批量,新手要引导和默认值。两者的界面几乎相反
数据量级管十条和管一万条,是两个产品。搜索、筛选、分页全都不一样
一个人还是一群人协作一旦进来,权限、通知、冲突处理全都要有,复杂度翻几倍
付钱的是不是本人自己掏钱看性价比,公司报销看合规和发票
Become irreplaceable to a small group first, then expand. The reverse almost never works — a product anyone can use is very hard to turn into the first choice of any one group later.
同样是切一刀,切的位置决定它有没有用 切对了:沿「使用场合」 在工位上 → 大屏、批量操作、快捷键 在路上  → 单手、可被打断、极简 两边的产品长得完全不同 —— 这一刀有用 切错了:沿「年龄」 25–35 岁 → 列表、筛选、导出 35–45 岁 → 列表、筛选、导出 产品做法一个字没变 —— 这一刀白切 值得切的五个维度 使用场合 熟练程度 数据量级 一个人还是一群 谁掏的钱 判断收得够不够窄有个土办法:你能不能列出五个符合描述的真人,并且知道去哪儿找到他们。 列不出来,就再收一层。
Become irreplaceable to a small group first, then expand. The reverse almost never works — a product anyone can use is very hard to turn into the first choice of any one group later.

What if the market is too small

宽的做法
面向所有自由职业者
潜在市场:很大
实际转化:每个人都觉得「差不多但不完全适合我」
大市场 × 极低转化 = 没有用户
窄的做法
面向用微信对接的设计师
潜在市场:小很多
实际转化:「这就是给我做的」
小市场 × 高转化 + 口碑 = 有用户
Become irreplaceable to a small group first, then expand. The reverse almost never works — a product anyone can use is very hard to turn into the first choice of any one group later.

A rough test for whether you have narrowed enough: can you name five real people who fit the description, and do you know where to find them? If you cannot, narrow another layer.