reddit · SaaS · 10-05 08:50 68 分 ← 返回列表

I ran 455 AI engine queries on my own company for three months. On the questions in my own category: 0 citations out of 196.

For three months I pointed the measurement at myself instead of a client. Six engines, 42 questions, 455 runs, July through late September. Three numbers came back and only one of them was comfortable. On my own name, 35 out of 35. Every engine, every time. On the category question, the ones somebody types when they already know what kind of thing they want to buy, 0 citations out of 196 measurements across 25 questions. Zero. On questions about the underlying problem, where nobody

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：AI搜索/答案引擎在商业意图的品类问题中不引用自家品牌，而现有AI可见性指标把品牌词、品类词和问题词混成一个分数，掩盖了真正的获客缺口。

目标用户：依赖AI搜索获客的SaaS创始人、增长营销负责人、SEO/GEO负责人和内容策略团队。

证据：On the category question, the ones somebody types when they already know what kind of thing they want to buy, 0 citations out of 196 measurements across 25 questions. Zero. ... The category questions are the ones carrying commercial intent, and they are held by something other than my competitors. Citations there went to video, to tool vendors running comparison pages, and to directories. Twenty of the twenty-five pulled from the same video platform. ... Has anyone split their own measurement this way, category questions on one side and problem questions on the other?

变现：订阅制SaaS，按监控的AI引擎数量、问题数、品牌数阶梯收费；用免费AI可见性审计报告获客，付费版提供品类问题引用率监控、竞品缺口分析和优化建议；可叠加内容/引用优化代运营服务。

MVP：2周内接入2-3个可编程AI引擎（如ChatGPT/Perplexity/Gemini API）并抓取Google AI Overview，让用户输入品牌、品类和问题集，自动归类为品牌词、品类词、问题词三类并跑prompt，输出引用率对比、引用来源域名分布和竞品/替代内容缺口，支持每周邮件报告。

竞品：Profound、Otterly.ai、Peec AI、Scrunch AI、Athena、Goodie AI

摘要：一位SaaS创始人用3个月、6个AI引擎、42个问题、455次查询测量自身，发现品牌词35/35被引用，但25个商业意图品类问题196次测量中0引用，问题类问题137次中38次被引用。品类引用主要流向视频平台、工具商对比页和目录站，现有单一AI可见性分数会掩盖这一缺口。发帖人询问是否应把品类问题和问题问题分开测量，这指向AI搜索意图分层监控工具机会。