reddit · SideProject · 09-24 18:50 51 分 ← 返回列表

I shipped the wrong thing once because feedback sat unread in my inbox. so I built FeedSense

solo dev here and this is mine, so take my enthusiasm with some salt. mid 2025 i was building another saas and i was too busy shipping features to listen to users. feedback came in through emails, form submissions, the occasional dm, and i kept telling myself i'd get to it later. i didn't. i built what i thought was right instead of what people actually needed. when i finally sat down and read months of it, the patterns were obvious. people had been telling me exactly what was wrong. i

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：用户反馈散落在邮件、表单、DM、Intercom、Zendesk、Slack 等多个渠道，创始人忙于开发没时间读，最终凭感觉做功能而做错方向。

目标用户：独立开发者、早期 SaaS 创始人、以及中小团队里负责收集与处理用户反馈的产品/客服负责人

证据：i built what i thought was right instead of what people actually needed. / none of them made decisions off a sentiment score. they decided off how many times the same thing came up / it's live with a free plan, paid from $49/mo

变现：SaaS 订阅制，按每月处理的消息量分层，免费版限约 200 条/月，付费从 $49/mo 起；可加团队席位、集成数量和自定义报表溢价。

MVP：2 周 MVP：只做 3 个输入源（CSV 上传、Slack Webhook、通用 Webhook）+ LLM 语义聚类去重，把同一诉求的不同说法归为一个 issue 并给出出现次数与环比增速，输出一个按"出现次数最多/增长最快"排序的待办清单，不做情感分析和花哨仪表盘。

竞品：Canny、Productboard、UserVoice、Enterpret、Kraftful、Cycle、Dovetail、Intercom AI、Zendesk AI

摘要：一位独立开发者在 2025 年中因没读散落各处的用户反馈而做错了产品方向，于是利用业余时间做了 FeedSense：接入 Intercom/Zendesk/Slack/表单/Webhook 等渠道，用 LLM 把同一诉求的不同说法聚类计数，输出优先级简报。目前已上线，免费版覆盖每月 200 条以下消息，付费从 49 美元/月起；他在三家公司的访谈中发现没人靠情感分做决策，只看同一问题被提及的次数。痛点真实但赛道极其拥挤，已有 Canny、Productboard、Enterpret 等成熟玩家。