Search

AI visibility

Monthly log

10 fixed prompts, same wording each month, run once on each engine on Fri 30 Oct. Next run Mon 30 Nov.

Geo grid Export log
ChatGPT1 of 10
Named for "water heater replacement in Austin" only
Gemini4 of 10
Named on prompts 1, 2, 4 and 5
Perplexity0 of 10
Not named on any prompt
One run per prompt is noisy. The same prompt can name different businesses the next day, so a single cell means little. Read the trend from month to month, per engine. There is no combined score because the engines rarely agree. Map pack rank still drives most calls, so the geo grid stays ahead of this in the queue.
Prompts, Fri 30 OctNamed = Brightwater Plumbing appears in the answer
#PromptChatGPTGeminiPerplexity
1best plumber in AustinNoNamedNo
2water heater replacement in AustinNamedNamedNo
3who can replace my water heater in Austin this weekNoNoNo
4emergency plumber in Austin open nowNoNamedNo
5slab leak repair company in AustinNoNamedNo
6how much does slab leak repair cost in AustinNoNoNo
7who does whole-house repipes in AustinNoNoNo
8drain cleaning in East AustinNoNoNo
9plumber in Austin with good reviewsNoNoNo
10tankless water heater installer in AustinNoNoNo
Named1 of 104 of 100 of 10
First month of the log. A trend appears after the Mon 30 Nov run.
How the log is kept
Prompts
Fixed wording, set on Fri 30 Oct. Changing one starts a new line in the log
Engines
ChatGPT, Gemini, Perplexity. Each scored on its own
Scored as
Named or not named; each cited source is stored with the answer
Separate from
Map pack rank. You can be in the pack and missing here, or the other way round
FoursquareNot claimed

Practitioners trace many of ChatGPT's local picks to Foursquare. Unverified, but claiming it is free. Use the same name, address and phone as the profile.

Not part of the 8 listings we check, so those counts stay 8 checked, 3 failing, 2 missing.