Purcell VenturesWhat did the check find?
- OpenStreetMap lists 60 small business websites inside the Tuscaloosa city limit. 9 of them (15.0%) had no working home page, against 15.4% statewide.
- Of the 47 working pages on the business's own domain, 21 (44.7%) carry Organization or LocalBusiness structured data and 9 (19.1%) publish an llms.txt file.
- 6 (12.8%) carry all three AI search signals I checked: business structured data, an llms.txt file, and no AI crawler blocked in robots.txt. Statewide the figure is 16.7%.
- 7 (14.9%) block at least one of five AI crawlers, about 4 times the statewide rate of 3.8%.
The Tuscaloosa figures cover every listed business domain in the city, so they count the whole OpenStreetMap list. The Alabama figures come from a random draw of 700 of the 2,102 business domains in the state. Both come from the same run of the Alabama study, with the same query, filters and script.
How does Tuscaloosa compare with Alabama?
The table lists every measure in the study, with the count and the base for each side. The last column gives a Fisher exact test of each gap. A value under 0.05 means a gap that large would turn up by chance less than one time in twenty. Only rows in the AI crawler group fall under it. One Tuscaloosa site moves a page figure by 2.1 points.
| Measure | Tuscaloosa (every listed site) | Alabama (random sample) | Gap in points | Fisher exact p |
|---|---|---|---|---|
| Does the home page work? | ||||
| Working home page (status 200) | 51 of 60 (85.0%) | 592 of 700 (84.6%) | +0.4 | 1.000 |
| No working home page | 9 of 60 (15.0%) | 108 of 700 (15.4%) | -0.4 | 1.000 |
| Web address no longer resolves | 2 of 60 (3.3%) | 57 of 700 (8.1%) | -4.8 | 0.309 |
| Working page ends on a different domain | 4 of 51 (7.8%) | 40 of 592 (6.8%) | +1.0 | 0.771 |
| Basics on the business's own page | ||||
| Ends on https | 44 of 47 (93.6%) | 503 of 552 (91.1%) | +2.5 | 0.788 |
| Plain http only, never https | 3 of 47 (6.4%) | 49 of 552 (8.9%) | -2.5 | 0.788 |
| Viewport tag for phones | 44 of 47 (93.6%) | 521 of 552 (94.4%) | -0.8 | 0.743 |
| Page title | 44 of 47 (93.6%) | 526 of 552 (95.3%) | -1.7 | 0.490 |
| Meta description | 34 of 47 (72.3%) | 405 of 552 (73.4%) | -1.1 | 0.865 |
| No meta description | 13 of 47 (27.7%) | 147 of 552 (26.6%) | +1.1 | 0.865 |
| All four: https, viewport, title, meta description | 31 of 47 (66.0%) | 379 of 552 (68.7%) | -2.7 | 0.744 |
| Structured data | ||||
| Any JSON-LD | 28 of 47 (59.6%) | 340 of 552 (61.6%) | -2.0 | 0.876 |
| Organization or LocalBusiness JSON-LD | 21 of 47 (44.7%) | 268 of 552 (48.6%) | -3.9 | 0.650 |
| AI crawlers and llms.txt | ||||
| robots.txt present | 45 of 47 (95.7%) | 510 of 552 (92.4%) | +3.3 | 0.565 |
| Blocks at least one of the five AI crawlers | 7 of 47 (14.9%) | 21 of 552 (3.8%) | +11.1 | 0.004 |
| Blocks all five | 1 of 47 (2.1%) | 4 of 552 (0.7%) | +1.4 | 0.336 |
| Blocks GPTBot | 6 of 47 (12.8%) | 17 of 552 (3.1%) | +9.7 | 0.006 |
| Blocks ClaudeBot | 7 of 47 (14.9%) | 17 of 552 (3.1%) | +11.8 | 0.001 |
| Blocks PerplexityBot | 2 of 47 (4.3%) | 6 of 552 (1.1%) | +3.2 | 0.125 |
| Blocks Google-Extended | 4 of 47 (8.5%) | 17 of 552 (3.1%) | +5.4 | 0.074 |
| Blocks CCBot | 6 of 47 (12.8%) | 20 of 552 (3.6%) | +9.2 | 0.012 |
| llms.txt file | 9 of 47 (19.1%) | 139 of 552 (25.2%) | -6.1 | 0.481 |
| All three: business JSON-LD, llms.txt, no AI crawler blocked | 6 of 47 (12.8%) | 92 of 552 (16.7%) | -3.9 | 0.681 |
| Speed from one computer in Tuscaloosa | ||||
| First byte took over 800 ms | 16 of 47 (34.0%) | 147 of 552 (26.6%) | +7.4 | 0.306 |
| HTML download took over 2 seconds | 4 of 47 (8.5%) | 45 of 552 (8.2%) | +0.3 | 0.787 |
| The OpenStreetMap listing itself | ||||
| Listing has opening hours | 36 of 60 (60.0%) | 400 of 700 (57.1%) | +2.9 | 0.686 |
| Listing has a phone number | 44 of 60 (73.3%) | 568 of 700 (81.1%) | -7.8 | 0.172 |
| Timings | ||||
| Median time to first byte | 614 ms (47 pages) | 617 ms (552 pages) | ||
| 75th percentile time to first byte | 972 ms (47 pages) | 828 ms (552 pages) | ||
| Median HTML download | 734 ms (47 pages) | 738 ms (552 pages) | ||
| Median HTML size | 141 KB (47 pages) | 140 KB (552 pages) | ||
| Area | Base n (listed web addresses) | Social page |
|---|---|---|
| Tuscaloosa | 63 | 3 (4.8%) |
| All of Alabama | 2,191 | 89 (4.1%) |
The Alabama row counts every listing in the state, before the random draw.
Why does AI crawler blocking stand out?
At the statewide rate, about 2 of the 47 Tuscaloosa pages would block an AI crawler. 7 do. The Fisher exact test puts the chance of a gap that large at 0.004, so chance alone is an unlikely cause. The count is still small, and 7 sites is a short list to explain a city with.
In 6 of the 7 Tuscaloosa pages that block a crawler, the blocking rule in robots.txt names the crawler. For the rest, the rule for all crawlers does the blocking. One page blocks all five. By crawler, the Tuscaloosa counts run ClaudeBot 7, GPTBot 6, CCBot 6, Google-Extended 4 and PerplexityBot 2. The check reads the rule and cannot tell who wrote it. Website platforms, hosts and SEO plugins add these rules for some owners.
| Area | Base n (own-domain pages) | Blocks GPTBot | Blocks ClaudeBot | Blocks PerplexityBot | Blocks Google-Extended | Blocks CCBot |
|---|---|---|---|---|---|---|
| Tuscaloosa | 47 | 6 (12.8%) | 7 (14.9%) | 2 (4.3%) | 4 (8.5%) | 6 (12.8%) |
| Alabama (random sample) | 552 | 17 (3.1%) | 17 (3.1%) | 6 (1.1%) | 17 (3.1%) | 20 (3.6%) |
What did the campus audit find?
On 20 August 2026 I counted every independent business within 2,500 m of the Quad, with or without a website. There were 121. Of those, 27 had a website a machine could reach, and 23 had opening hours recorded anywhere a machine can read them. The full count by kind of business is in 121 businesses near campus, and what a phone can find.
The November check starts from the listings that carry a website, so it skips the many businesses near campus that list none. Of the 60 Tuscaloosa domains, 36 (60.0%) have opening hours on the map listing. The two counts differ in area, date, kinds of business and test, so read them side by side and keep them apart.
What can Tuscaloosa owners fix?
- Run the free AI visibility check on your own site. It reads your robots.txt, llms.txt and home page and lists what to fix, including any rule that blocks an AI crawler.
- If your web address no longer works, claim your Google listing first, so customers can still find your phone number and hours.
- For structured data, llms.txt and crawler rules in one place, read AI websites for small businesses.
- To talk it through in person, AI consulting in Tuscaloosa covers how I work with businesses here.
Method
This page reads the Tuscaloosa frame of the Alabama study, which prints the full Overpass query and the exact test behind each check. The Tuscaloosa query ran on 9 November 2026 at 22:19 UTC against the OpenStreetMap boundary with Wikidata id Q79580, the Tuscaloosa city limit, and the statewide query ran at 22:05 UTC. Both ask for every listing tagged shop, craft or office, every restaurant, cafe, bar, fast food outlet, dentist, doctor's office, clinic and veterinarian, every hotel and every fitness centre, where the listing has a website or contact:website tag. Chains, government offices, schools, nonprofits and social media pages came out before the checks.
One script checked every domain once, from 22:32 to 22:41 UTC on 9 November 2026, from one computer in Tuscaloosa. It read the home page only, allowed 10 seconds per request and read robots.txt under RFC 9309 for GPTBot, ClaudeBot, PerplexityBot, Google-Extended and CCBot. The Fisher exact test on this page compares each Tuscaloosa count with the statewide count and its base.
Limits
- The Tuscaloosa base is small: 60 domains and 47 own-domain pages. One site moves a page figure by 2.1 points, and a gap of a few points means little.
- The statewide draw may hold some Tuscaloosa domains, so the two groups overlap a little. The overlap pulls the two sides together, which makes the test cautious.
- Each site got one try. A site that was down for a few minutes counts as not working, and some sites that block automated requests may show up as not working too.
- OpenStreetMap depends on volunteers. A business appears only if someone mapped it with a website, and some listings are out of date.
- robots.txt states a request to crawlers. It does not show firewall or hosting rules that turn AI crawlers away, which some hosts apply by default.
- The timings come from one computer in Tuscaloosa and cover the HTML only. They say nothing about PageSpeed or Core Web Vitals.
- The structured data check read JSON-LD only, so it missed microdata and RDFa.
- The campus audit used a different area, date, tag set and test, so its figures stay separate from the study figures.
Common questions
How do Tuscaloosa small business websites compare with the rest of Alabama?
On most checks they sit close to the Alabama statewide sample. 15.0% of Tuscaloosa sites had no working home page, against 15.4% statewide, and 44.7% carried business structured data, against 48.6%. The clear gap is AI crawler blocking: 7 of 47 Tuscaloosa pages block at least one AI crawler, against 21 of 552 in the statewide sample.
How many sites do the Tuscaloosa figures rest on?
60 business domains, every one that OpenStreetMap lists inside the Tuscaloosa city limit, and 47 working home pages that stayed on the business's own domain. One site moves a Tuscaloosa page figure by 2.1 points, so read small gaps with care.
Why would a small business site block AI crawlers?
The check reads the rule in robots.txt and cannot tell who wrote it. Some owners choose to block AI crawlers, and some website platforms, hosts and SEO plugins add the rules for them. A blocked crawler cannot read the page, so an AI assistant describes the business from other sources.
Can I check my own website the same way?
Yes. The free AI visibility check at purcellventures.co/ai-visibility-check reads your robots.txt, llms.txt and home page, including the title, meta description, structured data and https, and lists what to fix.
Data and license
Download the Alabama study CSV. Its tuscaloosa rows hold the count, base and percentage behind every Tuscaloosa study figure on this page, beside the statewide rows, with no domains or business names.
Listing data © OpenStreetMap contributors, available under the Open Database License. The aggregate figures on this page and in the CSV carry a Creative Commons Attribution 4.0 license. Credit Purcell Ventures and OpenStreetMap contributors when you use them.