Who is the best implant dentist in Los Angeles?
Which Los Angeles implant dentists do ChatGPT, Gemini, Perplexity and Google AI Overviews name when patients ask for a recommendation. This ranks what the engines say. It is not our recommendation.
asked, not named asked, and AI named them
We asked each engine 10 times. Total: 40 asks.
Often named · Issue 01
- 312
- dentists checked
- 10
- named at least once
- 10
- questions per engine
- 4
- AI engines
Often named
Same tier = no order. Listed A to Z.
Sometimes named
Same tier = no order. Listed A to Z.
Rarely named
Same tier = no order. Listed A to Z.
Protocol · Method v1
How this was measured
Between Oct 20 and Oct 31, 2026 we asked ChatGPT, Gemini, Perplexity and Google AI Overviews the same 10 recommendation questions, such as “Who is the best implant dentist in Los Angeles?” We pulled every person named out of the answers and matched them to licensed dentists. People whose ranges overlap share a tier; inside a tier, order is alphabetical. Changes are shown only when they exceed normal run-to-run variation. Results from different method versions can’t be compared.
Definitions
Named: the engine wrote the person into its answer. k of n: named in k of n answers. Tier: people whose likely ranges overlap.
Limits
Answers change from run to run; we repeat each question to measure that. Engines can confuse people with similar names; we resolve names to licensed dentists before counting.
Being named by an AI is not an endorsement, by us or by the engine.
Published by SoaRank. Practices that are SoaRank clients are marked; measurement is run separately from sales.
Download CSVFull method (v1) Checked, not yet named: 302
北京最值得关注的 AI 创业者有哪些?
当有人问「北京最值得关注的 AI 创业者」时,豆包、千问、DeepSeek 会点到谁。 这里记录的是 AI 怎么说,不代表我们的推荐。
问了,没有点到 名字出现在回答里
每个引擎问 10 次,共 30 次。
常被点名 · 第 1 期
- 186
- 位创业者测过
- 8
- 位至少被点名一次
- 10
- 道题 / 每个引擎
- 3
- 个 AI 引擎
常被点名
同一档不分先后,按姓氏拼音排列。
偶被点名
同一档不分先后,按姓氏拼音排列。
少被点名
同一档不分先后,按姓氏拼音排列。
实验方法 v1
怎么测的
2026 年 10 月 20 日至 31 日,我们向豆包、千问、DeepSeek 各问同样的 10 道推荐题,例如「北京最值得关注的 AI 创业者有哪些?」。从回答里抽出所有被点名的人,再逐一核对身份。区间重叠的人归同一档,档内按姓氏拼音排列。只有超出正常波动的变化才标出来。不同方法版本的结果不能直接比较。
名词
被点名:AI 在回答里写出了这个人。k / n:n 次回答里出现 k 次。档位:可能区间互相重叠的人归同一档。
局限
同一个问题每次回答都可能不同,所以每题重复问;AI 可能把同名的人混在一起,计数前我们先核对身份。
被提及不等于背书:既不代表我们推荐,也不代表 AI 推荐。
由 SoaRank 发布。是 SoaRank 客户的会标出来;测量和销售分开进行。