Lubbock / September 2026 / Descriptive study
A business can appear throughout the sampled Google results and rarely be named in an AI answer. Others appear across both. We examined 20 local industries to make those differences visible and checkable.
Google Organic, Google Maps, ChatGPT Search, Gemini, Claude and Perplexity. AI Overview panels are derived from Google Organic, not additional independent searches.
Editorial update: September 11, 2026. The lead stories now incorporate the reviewed Google-to-AI comparison using the same 6,000 observations. What changed · Original article and media brief
The study in 30 seconds
57%
About 57% (139 of 244) of tracked business website groups that appeared in Google's sampled top ten were rarely detected on at least one AI platform. This is low detection in sampled answers, not complete absence from AI.
Check the evidence50/50 vs 0/50
CoNetrix was named in 50/50 sampled ChatGPT answers and 0/50 Claude answers. Gorilla Law Firm showed the opposite contrast: 1/50 on ChatGPT and 40/50 on Claude. These are examples of platform differences, not a ranking of the assistants.
Check the evidence14/21 · 16/25
The pattern also appears in everyday local services: 14 of 21 tracked Google-visible restaurant website groups and 16 of 25 dentist groups were detected fewer than 5/50 times on at least one AI platform, even when known uncertain matches count as possible detections.
Check the evidenceRare means fewer than 10% of sampled answers on at least one of four AI platforms: fewer than 5/50 for a single-vertical group, or 10/100 when two verticals are pooled. It does not mean absent.
The 244 tracked business website groups each appeared at least once in a sampled Google top ten. They are not a census of Lubbock businesses or 244 consistently high-ranking businesses. Names and domains are different measurements; neither is a recommendation or business-quality score.
For residents, these contrasts show why one search surface may not show the same options as another. For businesses, they show why an organic ranking cannot stand in for an AI visibility measurement. This study does not measure consumer switching, clicks, leads or revenue.
Tracked business website groups, not all Lubbock businesses
Below 10% on at least one AI platform. Reviewed names; known uncertain matches count as possible detections.
50 sampled answers per business and platform
Name detections, not recommendations or a platform-quality ranking. These two businesses are selected examples.
Tracked Google-visible website groups within each sampled vertical
Fewer than 5/50 detections on at least one AI platform, including known uncertain matches as possible detections.
Explore the sampled question sets
Each vertical page contains the cross-platform visibility table, its five exact question pairs, AI Overview frequency and observed source-domain patterns. Compare the recorded names and domains, then inspect the counts by question.
Supporting research
The local lead stories above include the later name-review analysis. These original findings remain available as supporting evidence, including the role of directories and AI Overviews.
Finding 01
Gorilla Law Firm's domain appeared in the Google top ten in 50/50 sampled organic results. Its name matched 1/50 usable ChatGPT answers and 40/50 Claude answers.
Visibility needs to be checked by surface. This does not explain the cause or establish a business-wide absence.
Evidence: gorilla / organic · gorilla / chatgpt · gorilla / claude
Finding 02
CoNetrix matched 50/50 ChatGPT answers and 0/50 Claude answers in the managed IT prompt set.
A single AI platform cannot represent every platform. These configurations are not a consumer market-share sample.
Evidence: conetrix / chatgpt · conetrix / claude
Finding 03
Southwest Cleaning's domain appeared in 49/50 Google top tens. Its name matched 50/50 ChatGPT and 50/50 Perplexity answers.
This is an illustrative overlap case, not a category-winner claim or a recommendation rate.
Evidence: southwest / organic · southwest / chatgpt · southwest / perplexity
Finding 04
A Yelp domain appeared in 665/1000 sampled Google top tens and 809/1000 top twenties. Mobile and main domains are deduplicated within each result.
Discovery can take place inside a third-party result. Presence alone does not establish influence or an available placement.
Evidence: yelp.com / top10 · yelp.com / top20
Finding 05
Facebook domains appeared in 574/1000 sampled Google top tens and 785/1000 top twenties.
Social domains also appear in traditional search. These counts do not measure clicks, conversions or individual businesses.
Evidence: facebook.com / top10 · facebook.com / top20
Finding 06
Tripadvisor appeared in 96/100 Google top tens across the five hotel and five restaurant prompts, each repeated ten times.
The observed pattern is specific to these hospitality intents, not every local industry.
Evidence: tripadvisor / top10
Finding 07
Google returned AI Overview panels in 599/1000 organic observations. Foundation repair had 49/50, versus 2/50 for personal injury law. One returned panel had no answer text.
An AI Overview panel is not an additional independent search or proof that a business was cited.
Evidence: aio / all · aio / foundation-repair · aio / personal-injury-law
Finding 08
BBB was answer-linked in 251/1000 usable ChatGPT responses and 1/1000 Claude responses. Expertise appeared in 236/1000 and 130/1000, respectively.
These are observed answer links, not the platforms' complete retrieval histories or proof that a source caused an answer.
Evidence: bbb.org / chatgpt · bbb.org / claude · expertise.com / chatgpt · expertise.com / claude
Finding 09
Southwest Cleaning matched 10/10 Claude office-cleaning answers but 0/10 floor-care answers. It matched all ten floor-care answers on both ChatGPT and Perplexity.
A visibility gap in one sampled intent does not prove a service, content or capability deficiency.
Evidence: southwest / cal-commercial-cleaning-1 / claude · southwest / cal-commercial-cleaning-5 / claude · southwest / cal-commercial-cleaning-5 / chatgpt · southwest / cal-commercial-cleaning-5 / perplexity
Finding 10
Mean within-prompt set overlap was 98.0% for Maps and 88.7% for organic results. The calculation includes 4,500 Maps and 4,500 organic response pairs.
Set overlap is not rank-order stability. Cache behavior and statistical independence are not established.
Evidence: overlap / maps · overlap / organic
These are repeat observations of selected questions, not a random consumer survey, proof of ranking causes or an assessment of business quality. API and scraped platform configurations differ. Complete semantic recommendation rates and citywide business league tables remain withheld.
Study v2 includes six same-question ChatGPT retries and one revised personal-injury question on all six configurations. The original edition and outcomes are retained. See the version history and its effect on the findings.
Start with the story, then inspect its denominator, exact question and evidence record. No registration or sales form is required.
