Datasets:
Update card for 500-row sample
Browse files
README.md
CHANGED
|
@@ -20,7 +20,7 @@ configs:
|
|
| 20 |
data_files: ai_visibility_sample.csv
|
| 21 |
---
|
| 22 |
|
| 23 |
-
# AI Visibility Sample:
|
| 24 |
|
| 25 |
A free sample of the Engagemii AI visibility dataset. Each row is a real US business
|
| 26 |
website with a score from 0 to 10 for how findable and citable that site is to AI answer
|
|
@@ -28,7 +28,8 @@ engines such as ChatGPT, Claude, Gemini, and Perplexity, plus a record of which
|
|
| 28 |
crawlers have actually visited it.
|
| 29 |
|
| 30 |
The full dataset covers 34.7M+ websites worldwide, 15.1M of them verified US businesses.
|
| 31 |
-
This sample is
|
|
|
|
| 32 |
|
| 33 |
## Why this exists
|
| 34 |
|
|
@@ -44,7 +45,6 @@ problem with real data instead of anecdotes.
|
|
| 44 |
| `companyName` | Business name |
|
| 45 |
| `domain` | Registered domain |
|
| 46 |
| `websiteUrl` | Resolved site URL |
|
| 47 |
-
| `scoredAt` | Date this row was scored |
|
| 48 |
| `aeoScore` | AI visibility score, 0 to 10 |
|
| 49 |
| `aiBotHitsTotal` | Observed AI crawler requests to this site |
|
| 50 |
| `aiBotsSeen` | Which AI crawlers were observed, by name |
|
|
@@ -58,11 +58,12 @@ problem with real data instead of anecdotes.
|
|
| 58 |
|
| 59 |
## What is in this sample
|
| 60 |
|
| 61 |
-
In these
|
| 62 |
|
| 63 |
-
- Scores run from 0.
|
| 64 |
-
-
|
| 65 |
- ClaudeBot is the most frequently observed crawler in this sample, followed by GPTBot.
|
|
|
|
| 66 |
|
| 67 |
The score is a 0 to 10 measure built from on-page and off-page signals. The component
|
| 68 |
weights are not published.
|
|
@@ -91,10 +92,10 @@ separately, by checking the answers themselves.
|
|
| 91 |
|
| 92 |
```bibtex
|
| 93 |
@misc{engagemii_ai_visibility_sample,
|
| 94 |
-
title = {AI Visibility Sample:
|
| 95 |
author = {Pellitteri, Gregory},
|
| 96 |
year = {2026},
|
| 97 |
-
url = {https://huggingface.co/datasets/
|
| 98 |
}
|
| 99 |
```
|
| 100 |
|
|
|
|
| 20 |
data_files: ai_visibility_sample.csv
|
| 21 |
---
|
| 22 |
|
| 23 |
+
# AI Visibility Sample: 500 US business websites
|
| 24 |
|
| 25 |
A free sample of the Engagemii AI visibility dataset. Each row is a real US business
|
| 26 |
website with a score from 0 to 10 for how findable and citable that site is to AI answer
|
|
|
|
| 28 |
crawlers have actually visited it.
|
| 29 |
|
| 30 |
The full dataset covers 34.7M+ websites worldwide, 15.1M of them verified US businesses.
|
| 31 |
+
This sample is 500 rows drawn from the US set, spread across 20 business categories and
|
| 32 |
+
48 states.
|
| 33 |
|
| 34 |
## Why this exists
|
| 35 |
|
|
|
|
| 45 |
| `companyName` | Business name |
|
| 46 |
| `domain` | Registered domain |
|
| 47 |
| `websiteUrl` | Resolved site URL |
|
|
|
|
| 48 |
| `aeoScore` | AI visibility score, 0 to 10 |
|
| 49 |
| `aiBotHitsTotal` | Observed AI crawler requests to this site |
|
| 50 |
| `aiBotsSeen` | Which AI crawlers were observed, by name |
|
|
|
|
| 58 |
|
| 59 |
## What is in this sample
|
| 60 |
|
| 61 |
+
In these 500 rows:
|
| 62 |
|
| 63 |
+
- Scores run from 0.6 to 8.1. The median is 4.7. No site in the sample reaches 9.
|
| 64 |
+
- 304 of 500 sites had at least one identified AI crawler visit recorded.
|
| 65 |
- ClaudeBot is the most frequently observed crawler in this sample, followed by GPTBot.
|
| 66 |
+
- Every domain is unique. Records are spread across 20 business categories and 48 states.
|
| 67 |
|
| 68 |
The score is a 0 to 10 measure built from on-page and off-page signals. The component
|
| 69 |
weights are not published.
|
|
|
|
| 92 |
|
| 93 |
```bibtex
|
| 94 |
@misc{engagemii_ai_visibility_sample,
|
| 95 |
+
title = {AI Visibility Sample: 500 US business websites},
|
| 96 |
author = {Pellitteri, Gregory},
|
| 97 |
year = {2026},
|
| 98 |
+
url = {https://huggingface.co/datasets/Engagemii/ai-visibility-sample}
|
| 99 |
}
|
| 100 |
```
|
| 101 |
|