Reviewed by Irfan Ahmad
Ask ChatGPT which roofer to call in Dallas, TX and you get a name. Ask Perplexity and you may get a different one. Ask again tomorrow and the answer can change. NameBeam, a small Wyoming company, decided to write all of it down.
Since September 1, 2026, NameBeam has asked four AI engines (ChatGPT, Perplexity, Gemini and Claude) the same buyer questions in 46 US market-and-trade pairs, once a day. Every answer is saved, hashed and recorded. The counts are in a public file and the raw answers sit behind each market.
In the window from September 1st - October 11th, 36 of the 46 markets had two or more engines naming at least one business. In only 14 of those 36 did two or more engines pick the same business as their most named answer. The other 22 markets split. A buyer in those cities gets a different first name depending on which app they open.
Roofing shows the split most clearly… Take residential roofing in three cities. In Dallas, Perplexity named Arrington Roofing on 27 of its 37 answer days (https://ift.tt/av9N6h7). In Denver it named Gates Enterprises on 35 of 37. In Miami it named Extreme Roofing on 26 of 37.
ChatGPT and Claude named none of those 3 companies on any of their 40 answer days. Gemini, asked fewer times because of its quota, named them on 0 of 8, 0 of 9 and 0 of 10 days. Same city, same question, same six weeks. One engine returned a steady name. Two returned with nothing for it at all.
Across all 46 markets, Perplexity named at least one business in 37 of them. Gemini did in 32, ChatGPT in 24 and Claude in 9.
When an engine did name a business, how steady was it? Perplexity's most-named business showed up on 56 percent of its answer days on average. Gemini's was 72 percent across the 40 markets where it answered. ChatGPT's was 20 percent and Claude's 7 percent.
The method is plain. A business counts for an engine on any day the saved answer names it in full. Each engine is measured against its own answer days, so a quota-limited engine is not punished for being asked less often.
What the record does not claim
The record does not say why an engine picks a company. It does not say that a website change or a review caused a name to appear. It records what was said, on which day, by which engine, and keeps the text so the count can be rerun. Namebeam sells that record to agencies and platforms that serve local businesses, not to the businesses being measured.
The practical point is simple. "Does AI recommend us?" is not one question anymore. It is four, and they can have four different answers. A dated record is the only way to know which.
About the author
Terry Carter is the founder of Carter Enterprise LLC in Sheridan, Wyoming. He runs Namebeam, a public record of how AI engines name local businesses. Every figure he publishes carries a date and a link readers can open.
Read next:
• Scientists Outrank Government Officials but Not Business Leaders, as 56% Question Their Independence, Bias or Relevance, Study Finds
• AI Models Reported 88% Confidence but Were Correct 79% of the Time, Study Finds
by Guest Contributor via Digital Information World
No comments:
Post a Comment