About a minute. We are opening the pages a buyer actually lands on, not just the homepage.
Anything we could not read is marked as such. It is never reported as missing.
46 checks across 17 pages. Each row says what it is worth, not just what it is.
Out of everything below, ranked by what it costs them.
Some paths are kept from every crawler (Disallow: /sample-*, /blog/sample-*, /_hcms/preview/). That is normal housekeeping, not a block.
Read robots.txt and checked all 10 AI crawlers by name · open
The one file that tells assistants which pages matter and what the company does, in their own words. About ten minutes to write, and almost none of their competitors have one. A nice-to-have, never a headline.
Read the sitemap, following the index where there was one · open
Checked each page for a 'noindex' instruction in its code and its headers
Most AI assistants read a page without running any JavaScript. Whatever is not there before the browser does its work is invisible to them.
Loaded each page the way an AI crawler does, with JavaScript switched off
Loaded each variation of the address and followed where it went
Read the structured data on 17 pages
Found in the structured data on 17 pages · open
Found in the structured data on 17 pages · open
Found in the structured data on 17 pages · open
You have pages for the towns you serve, but the site's structured data, the hidden labels a machine reads, never names them. So an assistant asked about a nearby town has nothing to go on.
Your reviews are on the site for people to read, but the review count and rating are not in the structured data an assistant reads, so it has no number to use when it chooses who to recommend.
Your FAQ page holds 32 questions, but none is marked up as a question and answer, which is what lets an assistant lift one straight out.
Your services are on the site for people to read, but not in the hidden labels a machine reads, so an assistant asked 'who does X' has to work it out from the prose.
Read the title of each page
Over about 60 characters a title gets cut off in search results; under 30 it usually says too little to be chosen. Neither is fatal, and both are quick to fix.
Compared the titles of every page we read
Read the description of each page
Over about 160 characters a description gets cut off; under 70 it rarely says enough to earn the click.
Counted the main headings on each page
Headings are the outline a machine reads a page by. Jumping from a main heading to a sub-sub-heading breaks the outline, and the section below it loses its place. Every page does it, so it is the site's template rather than any one page.
Looked for headings phrased the way a person would ask
Searched the 701 addresses in the sitemap and the 43 links on the homepage for one page of each kind, read each one to check it is what its name says, then the rest of the navigation
The people are on the page but not in structured data, so a machine reads them as text rather than as who runs the company. Worth adding.
Read /about-us · open
Read /contact · open
Read /charleston-property-management · open
Read /cities-served, /property-management-services-in-charleston-keep-you-successful, /daniel-island-property-management-company-services-in-daniel-island-sc, /goose-creek-rental-property-management-companies-services-in-goose-creek, /hanahan-property-management-services-help-investors-grow · open
An assistant matches the questions people ask against the questions a page answers. A question folded into an accordion or plain text is far harder to find than a heading, and FAQ structured data is what lets the answer be lifted straight out. There is also a knowledge base of 32 answers under /en/knowledge.
Read /owner-faqs · open
Read /pricing · open
Looked for a reviews page and for testimonials in the text · open
Found the blog at /blog · open
Read the dates on the newest post and in the sitemap · open
An article signed by a company name is an anonymous article to a reader and to an assistant. Put a real person's name and role on each post — it is the cheapest trust signal there is.
Read the newest post at /blog/can-i-exclusively-rent-to-females-charleston-property-management-weighs-in · open
Counted the words on each page, contact page aside
The canonical tag tells a crawler which address is the real one when a page can be reached several ways. Without it, the same page can be counted twice and credited to neither.
Checked each page for the viewport instruction
Read the language attribute on each page
Checked each image for a description
Followed the links between the pages we read
Checked each page for scripts, images or styles loaded over plain http
Measured the size of each page as sent
Timed one request to the homepage from our server
Read the Open Graph tags on the homepage
Followed a sample of the internal links on the pages we read
Google is still measuring. This row fills in by itself, usually within a minute or two.
Requested an address that does not exist and read the status
Asks ChatGPT, Claude, Perplexity and Google Gemini the questions their customers would type about charlestonpc.com, twice each and a third time when the two answers disagree, and comes back with every answer and every source.
The questions are written first and shown to you. Change, remove or add any of them, then press Ask — nothing goes to an assistant until you do. The asking takes a few minutes, and nothing is stored yet, so keep the tab open until it finishes.
Read live from charlestonpc.com on 9/25/2026, 5:47:09 PM. Nothing is stored yet.