Notes · 2026 Report
We checked 25 West Georgia business websites. Here's what we found.
In July 2026 we ran technical checks on 25 real business websites in Villa Rica, Carrollton, Douglasville, and Temple — dentists, chiropractors, law firms, HVAC and plumbing companies, a salon, an auto shop, and a restaurant. Nobody gets named. The numbers speak for themselves.
The short version
- 18 of 25 (72%) have nothing telling AI search tools who they are — no llms.txt file at all.
- 3 of 25 actively block AI crawlers like ChatGPT's and Perplexity's in their robots.txt — almost certainly without knowing it. Those businesses can't be found or recommended by AI search.
- 1 in 5 has a copyright footer from 2024 or earlier. One site still says ©2013.
- 4 of 25 (16%) have zero structured data — the machine-readable info Google and AI tools use to understand a business.
- The good news: all 25 have working HTTPS. The padlock problem is solved around here. The AI-visibility problem is not.
72% of the West Georgia business sites we checked are invisible-by-default to AI search — and 3 of them are actively telling it to go away.
Who we checked and how
We picked 25 established local businesses across the categories owners around here actually run: 8 dental practices, 7 chiropractic clinics, 4 law firms, 3 home-service companies (HVAC and plumbing), plus a salon, an auto shop, and a restaurant. All in the Villa Rica / Carrollton / Douglasville / Temple corridor of west metro Atlanta.
The checks were static and technical, run on July 14, 2026: we fetched each homepage the way a browser does and looked at HTTPS, the mobile viewport tag, titles and meta descriptions, headings, structured data (JSON-LD), robots.txt rules for AI crawlers, sitemap.xml, llms.txt, raw page weight, copyright dates, and tap-to-call links. This was not a full speed-lab performance run, and it's not a design review — it's the plumbing that decides whether search engines and AI tools can find, read, and recommend a business.
The full numbers
| Check | Result | Why it matters |
|---|---|---|
| Working HTTPS + secure redirect | 25 / 25 pass | Browsers flag non-HTTPS sites as "Not secure." Everyone passed — genuinely good news. |
| No llms.txt (AI-search summary file) | 18 / 25 fail | AI tools get no clean summary of the business — they guess from whatever they can scrape. |
| robots.txt blocks AI crawlers | 3 / 25 fail | These sites tell ChatGPT, Claude, and Perplexity crawlers to stay out. AI search can't cite what it can't read. |
| Zero structured data (JSON-LD) | 4 / 25 fail | No machine-readable name, hours, location, or services for Google and AI to rely on. |
| No sitemap.xml | 3 / 25 fail | Search engines have to discover pages by luck instead of a map. |
| Missing meta description | 3 / 25 fail | Google writes its own (often worse) snippet for your listing. |
| Homepage HTML over 200 KB | 11 / 25 | Heavy code before a single photo loads. The heaviest homepage we saw was 602 KB of raw HTML. |
| Copyright footer 2024 or older | 5 / 25 | Signals neglect to visitors and to Google. Oldest: ©2013. |
| Image-heavy page, no lazy loading | 4 / 25 | Every photo loads at once — slower first impression on phones. |
| No tap-to-call phone link | 3 / 25 fail | On a phone, visitors must copy-paste your number to call you. For a local business, that's the whole game. |
| Not mobile-ready (no viewport tag) | 1 / 25 fail | The site renders as a shrunken desktop page on every phone. |
Two additional findings didn't fit the table: two sites' firewalls returned "403 Forbidden" to anything that isn't a mainstream browser — which also shuts out legitimate search and AI crawlers — and 7 of the 25 sites do have an llms.txt file, but every one of them appears to be auto-generated by an SEO plugin, so the owners likely don't know it exists.
What this actually means if you own a business here
The last decade's website checklist — get online, get HTTPS, look decent on a phone — is mostly done in west Georgia. The 2026 gap is different: people increasingly ask ChatGPT, Perplexity, and Google's AI results who to call, and most local sites give those tools nothing to work with. When an AI can't read your site, it recommends the competitor it can read.
The frustrating part is how fixable this is. Almost everything in the table above is an afternoon of work for someone who knows where the files go: an llms.txt file, a corrected robots.txt, structured data that says what you do and where, a sitemap, real meta descriptions, a current footer.
The 5-minute self-check
- Open your site on your phone. Can you tap the phone number to call?
- Visit
yourdomain.com/llms.txt— do you get a page describing your business, or an error? - Visit
yourdomain.com/robots.txt— if you see "GPTBot" or "PerplexityBot" next to the word "Disallow," AI search is locked out. - Ask ChatGPT or Perplexity: "Who is [your business name] in [your town]?" — is the answer accurate, thin, or wrong?
- Look at your own footer. What year does it say?
Want your site's individual results? If your business is in the Villa Rica–Carrollton–Douglasville area, email info@ajivphotography.com and we'll run the same checks on your site and send you the scorecard, free — whether or not you were one of the 25, and with no obligation to buy anything.
Questions owners have asked us
Was my business one of the 25?
If you run a dental, chiropractic, law, or home-service business in the area, it's possible. Email us and we'll tell you — and send your site's results either way.
What is llms.txt?
A plain-text file on your website that gives AI tools a clean summary of who you are, what you do, and which pages matter — so AI search describes your business accurately instead of guessing. It's one signal among several, not a magic switch, but it's cheap and most of your competitors don't have it.
My site was built recently — am I fine?
Not automatically. Several modern-looking sites in this audit blocked AI crawlers via template defaults. Recent build usually means the mobile basics are covered; AI-search readiness is a separate checklist.
Will you do this audit again?
Yes — this is the first edition of an annual report. We'll re-run the same checks on the same categories in 2027 and publish whether west Georgia's numbers moved.
Method notes: checks run July 14, 2026 against each site's homepage and standard root files, using a standard browser user-agent. Aggregate results only; no individual businesses are named, and per-site data isn't published. "Blocks AI crawlers" means robots.txt disallows GPTBot, ClaudeBot, PerplexityBot, or Google-Extended from the homepage.