Press, September 2026 AI Visibility Corpus

What was measured

Every day in September 2026, NameBeam asked the same buyer-intent questions (“best [service] in [city]”-style local queries) to ChatGPT, Claude, Perplexity, and Gemini, with Google SERP as a baseline, across 22 local US markets. Every answer was saved with a date. The metric: distinct local businesses each engine named per market, summed across September 1 to 28.

Key findings

Engine	Businesses named per niche-market
Gemini	21.2
OpenAI (ChatGPT)	18.7
Perplexity	17.8
Google SERP (baseline)	13.7
Anthropic (Claude)	1.5

Normalized per niche-market, raw totals mislead because markets track different niche counts (Pleasanton 8, Los Angeles 1). We published the correction against our own first headline: see the recap.

Claude is the outlier. It named zero businesses in 7 of 9 local markets where it was asked about roofers, HVAC techs or lawyers (Sept 1 to 28). Denver and Liberty, MO were the exceptions. In Dallas it named 0 while Gemini named 66 and Perplexity named 41.

The engines disagree market to market. Gemini named 314 distinct businesses in Pleasanton, CA, and 0 in Los Angeles and Maryland. Same questions, same engines, different city, different answer set.

OpenAI named businesses in every one of the 22 markets, from 262 on the nationwide query set (one of the 22) down to 6 in Tyler, TX.

The story inside the data

Arrington Roofing, Dallas TX. Perplexity named this residential roofer on 24 of 28 days in September, more than any other Dallas business, straight through the final run day. Google's most-cited list for Dallas doesn't name them once. The company winning the AI era is invisible in the old phone book, and the old phone book is where every local SEO dollar still goes. Read the Dallas report.

All Star Movers, Pleasanton CA. Named by Perplexity on 25 of 28 days, the most-named business in the entire corpus. Google cited its site on 10 days too, but Google's top-cited Pleasanton business was a salon Perplexity barely notices. Two phone books, two different champions. Read the Pleasanton report.

Journalists: the corpus is the credibility, the businesses are the hook. Interview, per-market data, and the full 179-business Perplexity-only list on request.

Methodology

Corpus: September 1 to 28, 2026 (not every engine answered with names every day), 22 local US markets (21 metros + 1 nationwide query set).

Systems tested: ChatGPT (OpenAI), Claude (Anthropic), Perplexity, Gemini (Google), plus Google SERP as a non-AI baseline.

Metric: distinct local businesses named per engine per market, summed across run days. Answers change over time; no rankings are promised.

Transparency: per-market reports are public and dated. NameBeam tests on itself first, in public, at proof.namebeam.ai.

Press releases

Palm Beach Tan Is the Most-Named Tanning Brand in Our September Corpus, 88 Namings in 28 Days, October 1, 2026

Assets & links

All 24 September 2026 market reports, per-market dated findings
September 2026 recap, the engine-gap data story with heatmap
The AI-Named badge, free embeddable badge for named businesses
proof.namebeam.ai, NameBeam's public self-test ledger

About NameBeam

NameBeam, operated by Carter Enterprise LLC, runs a free 5-engine AI visibility check for local businesses, ChatGPT, Claude, Perplexity, Gemini, and Google AI, with dated transcripts. Motto: “Every claim owes the future a receipt.”

Press contact: [email protected]
Happy to share methodology detail, per-market data, or the full corpus notes on request.