Readiness study · New York City · 2026-08-31

828 of 1,106 New York City businesses publish nothing an AI agent can read.

We read 1,494 business websites in New York City and the area around it on 2026-08-31, as part of a 6,284-business study across 10 industries and 10 areas. 1,106 of them gave us an answer we could assess. 75% of those publish no /llms.txt — the file that exists for one purpose, telling an AI assistant what a business is. Of the 278 that do publish one, almost none of their owners wrote it.

1,494New York City businesses read
828 of 1,106assessable businesses publish no file for agents (75%)
791 of 996publish none of the four agent-facing documents (79%)
55block an AI crawler outright

What “New York City” means here

Businesses were found in OpenStreetMap’s open data, filtered to those publishing a website of their own, inside the bounding box 40.5, -74.3, 41, -73.7New York City and the area around it. That is the query, not a marketing description of a region, and anyone can re-run it. A business outside that box is not in this page, and its absence is not a finding about it.

388 of the 1,494 could not be assessed for this measure — a server that answered 503 is a server that failed to answer, not a business that publishes nothing — so they are excluded from the percentage rather than counted as absent. Every share on this page is taken over the 1,106 that could be decided.

We checked what a public HTTP GET returns: /llms.txt, three /.well-known/ documents, robots.txt and a sitemap, following each site’s own robots.txt. Nothing was submitted. No page was altered. No business was contacted.

By industry, inside New York City

The same ten industries as the national study, narrowed to this area. A share is printed only where at least 30 businesses in that industry gave an assessable answer here. Below that the counts stand on their own: a percentage over a handful of businesses tells a reader more than we know.

IndustryReadAssessablePublish a file for agentsShare
Professional services2521723722%
Self-care2501835430%
Hospitality2481834625%
Clinical2201773017%
Wellness1971463121%
Trade suppliers1401023231%
Building trades119992525%
Galleries34188
Antique dealers272011
Events & bridal764

Who wrote the files that do exist

Of the New York City businesses whose /llms.txt we could re-read, these name their own author — and the authors are software. This is the finding the national study leads on, and it holds here: presence is arriving as a platform default. Nobody at these businesses decided it.

Wrote the fileFiles in New York City
Wix73
Shopify51
Yoast SEO23
All in One SEO19
Rank Math5
WordPress (unspecified plugin)2
other self-declared generator1

A platform-generated file says what the platform decided to say. It is assembled from the pages and products already on the site. It does not say what the business will quote a price on, which services it actually performs, or the area it will travel to. The business is the subject of the description rather than its author.

What we withdrew, including about businesses here

This study published a column reporting whether each homepage carried organisation data. On 2026-09-03 we proved our own check wrong — it reported the data missing on homepages that publish it — and withdrew the column for every business in the study, including the ones named on this page. We are not publishing a replacement figure, because we know the old one was wrong and we do not know the right one. What happened, in full.

If your business is on this page and something is wrong, we would rather be told than be quoted. Tell us and we will correct it in place with a note saying what changed.

Look up a New York City business

Every one of the 1,494 businesses read here is listed with the facts observed about it: which documents were present, which absent, which could not be read, and which we have withdrawn. No business is scored, graded or ranked. We report what a URL returned on 2026-08-31.

Download the complete dataset as CSV or as JSON — all 6,284 businesses, every area, free to reuse with attribution.

Check your own site instead of reading about it

This page is a reading taken on 2026-08-31. Your site today is a different question, and you can settle it in about eight seconds. Our free check reports the URL and the HTTP status behind every line, so you can open the same address and see the same thing, and it returns unknown rather than missing when it could not read something.

Check what an AI agent can read about your site

The other areas we read

The same checks, the same day, in nine other places. The national picture is in the full study.

Cite this

River Cade Concepts, Agent presence among 6,284 small businesses, 2026-08-31. New York City subset: 1,494 businesses. https://rivercadeconcepts.com/reports/agent-presence/newyork

Reuse the data and the findings freely, with attribution. If you quote a percentage, quote the sample it was taken over with it — that is the whole discipline this study is built on, and we have already published one correction for breaking it ourselves.