We Audited 45 Tampa Real Estate Websites. 1 in 4 Blocks ChatGPT.

Short answer: We ran every Tampa real estate agency listed on Google Maps through an AI-visibility
audit. 24% actively block GPTBot (the crawler that feeds ChatGPT) with a hard 403. 42% publish no
structured data at all, so AI can't tell what the business even is. Put together, 44% are effectively
invisible to AI answer engines. When a buyer moving to Tampa asks ChatGPT "who's a good agent here,"
nearly half the market can't be recommended, no matter how good they are.

Original research by Specularis, July 2026. Raw data and methodology are published below.

What did the study find?

We audited 45 Tampa real estate websites (neutral market sample, see methodology).

Two separate failures are stacking on top of each other.

How many Tampa real estate websites block ChatGPT?

Eleven of forty-five, or 24%, return a hard 403 Forbidden when the request comes from OpenAI's
GPTBot user-agent. Five of those blocks come from Cloudflare, which serves a JavaScript challenge page
("Just a moment…") that a crawler can never solve.

Almost none of these firms chose this. A 403 to GPTBot is usually a default setting in a security product,
switched on by a web host or a well-meaning IT contractor. The business never made a decision. It just
quietly disappeared from AI answers.

Does blocking AI crawlers actually protect your website?

No. This is the most expensive misconception in the data.

robots.txt is a polite request, not a wall. The bots that actually threaten your site, like scrapers,
credential stuffers, and DDoS traffic, ignore robots.txt entirely. They don't announce themselves and
they don't read your rules. So disallowing GPTBot buys you exactly zero protection from attackers.

What it does buy you is removal from the answer when a buyer asks an AI for a recommendation.

Real protection lives at the firewall/WAF layer: rate limiting, bot-fight rules, IP reputation. That stays
fully intact whether or not you let the well-behaved AI crawlers through. Allowing GPTBot and
PerplexityBot does not make you less secure. These are documented, rate-limited, identifiable crawlers
from public companies.

Why does structured data matter so much here?

Getting crawled isn't enough. AI also has to understand what you are.

42% of the sites we audited publish no structured data whatsoever. No JSON-LD, nothing. And only
11% use RealEstateAgent schema, the machine-readable label that says "this is a real estate
professional, in this city, with these credentials."

Without it, an AI engine looking at the page sees some text and photos of houses. It can't confidently say
"this is a real estate agent serving Tampa." So when it's asked to name three agents, it reaches for the
ones it can verify: clean schema, consistent profiles, third-party corroboration.

You don't lose to a better agent. You lose to a more legible one.

What does this mean for a Tampa real estate agent?

The buyer behavior already shifted. Someone relocating to Tampa increasingly opens ChatGPT or Perplexity
and asks a question instead of scrolling a page of blue links. The AI returns a short list of names.

If you're not on that list, you don't get ranked lower. You're absent. There's no page two of an AI
answer. And you never see the lead, so you don't even know it happened.

The good news buried in this data: almost nobody in this market has claimed it yet. Only 11% have
proper schema. Only 13% have an llms.txt. This isn't a crowded race. It's an empty one, for now.

We ran the identical audit down in Miami and found the same story, only worse: 63% of Miami real estate
sites are invisible to AI. Two markets, one pattern. This
is an industry default nobody turned off, not a Tampa quirk.

How we ran the study (methodology)

We published the full method so anyone can verify or repeat it.

1. Sample. Pulled every "real estate agency" listing in Tampa, FL from Google Maps that has a website
   (July 2026), then deduplicated to unique root domains. No pre-filtering for AI visibility. This is a
   neutral market sample, not a list selected to look bad.

2. Exclusions. We removed 3 sites that were unreachable or returned 404 to every user agent,
   including a normal browser. Those sites are broken or offline. Counting them as "blocking AI" would have
   overstated the finding.

3. Re-testing. One site initially returned a 429 (rate limit). On re-test it returned 200, so we
   reclassified it as allowed. Transient errors are not blocks.

4. Crawler test. Requested each homepage using OpenAI's GPTBot user-agent. A hard 403 counts as
   blocked, and so does a Cloudflare JavaScript challenge.

5. Identity test. Fetched each homepage with a standard browser user-agent and counted
   application/ld+json structured-data blocks, checking for RealEstateAgent, LocalBusiness, and
   Organization types.

6. Emerging signals. Checked robots.txt for any AI user-agent directives, and /llms.txt, making
   sure it returned real text rather than a soft-404 HTML page.

We report aggregates only. We don't name individual businesses.

How can I check my own website?

Three checks you can run yourself in about five minutes:

1. Crawler access. Open yoursite.com/robots.txt and look for Disallow rules under GPTBot,
   ClaudeBot, PerplexityBot, or OAI-SearchBot. Then check with your host or Cloudflare whether an
   "AI scrapers" blocking rule is switched on. That block won't appear in robots.txt at all.

2. Structured data. View your homepage source and search for application/ld+json. Nothing there
   means AI has no reliable way to identify you.

3. The real test. Ask ChatGPT and Perplexity, "who's a good real estate agent in Tampa?" a few times
   and see whether your name ever appears.

Or run the free AI visibility audit. It checks all of the above
plus off-site signals, scores your site 0 to 100, and emails you the exact fixes. Free, no credit card.

For the numbers across every market we've studied, see our AI visibility statistics.

FAQ

How many Tampa real estate websites block ChatGPT?
24%. Eleven of the 45 sites we audited return a hard 403 to GPTBot, the crawler behind ChatGPT. Five of
those blocks come from Cloudflare.

Does blocking GPTBot protect my site from attacks?
No. Malicious bots ignore robots.txt and crawler rules entirely. Blocking AI crawlers provides no security
benefit. It only removes you from AI-generated answers. Real protection comes from your firewall/WAF, which
keeps working whether or not AI crawlers are allowed.

What percentage of real estate sites have proper schema?
Only 11% of the sites we audited use RealEstateAgent schema, and 42% publish no structured data at all,
so AI cannot reliably determine what the business is.

What is llms.txt and do agents need it?
llms.txt is a plain-text file at your site root that tells AI what your site is about. It's an emerging
standard, and just 13% of the sites we audited had one. It won't fix a blocked crawler, but it helps AI
understand you faster.

Is this study repeatable?
Yes. The method is simple and reproducible: sample from Google Maps, request each homepage as GPTBot, and
check for structured data. Our raw data is published alongside this article.