For most of my consulting career, a technical SEO engagement started the same way: crawl the site, export the findings, sort the spreadsheet, and write up what matters. The crawl was always the easy part. The hard part was that the report described the site on one particular day, and the site did not stop changing when the crawler did.

The worst problems I have seen were never there on audit day. A release added noindex to a template. A front-end refactor moved the product copy behind JavaScript. A migration redirected a whole section to the homepage. Each one shipped between audits, and each one was found by a traffic chart, weeks later.

Starting today, VitalSentinel has Site Health Monitoring: a technical SEO crawl that never stops. It is live on every plan, including Free.

An Audit Is a Photo

A scheduled audit is a photo of your site. It is a useful photo, but it answers "what was wrong on the 1st?" when the question you actually have is "what broke since Tuesday's deploy?"

Running the audit more often does not really fix that. A full crawl every day is heavy on your server and produces the same list of hundreds of findings every morning, most of which were already there last month. What you want is to be told when something changes for the worse, and to be told which change is worth your afternoon.

That is what Site Health is built around.

How It Works

Instead of crawling everything at once, our crawler checks your site continuously, one page at a time: one page a minute per domain on paid plans, and one every five minutes on Free.

It picks the pages that matter first. Your homepage, your top pages in Search Console and any pages you pin come first, then the pages in your sitemap, and random picks keep every part of the site represented. Each page in the selection comes around about once a week, and pinned pages about every day. On paid plans the selection holds up to 10,000 pages per domain, and on Free up to 2,500. You can pin up to 50 pages on every plan.

You decide what goes in. In Settings you set how many pages are checked, whether the pages with the most Google clicks, the most visits or the ones in your sitemap come first, and rules to only, always or never check a path.

Before and After JavaScript

Every check reads two versions of the page: the HTML your server sends, and the page after JavaScript runs.

This matters more every year. Google reads the HTML first and renders the page later. Many AI crawlers never run JavaScript at all. A site can look perfect in a browser while the HTML a crawler reads first is missing its text, its internal links or its structured data. Site Health compares the two versions, so Page text missing from HTML, Page tags rewritten by JavaScript, Noindex added by JavaScript and Page renders blank are issues it can raise, not things you discover in a rendering test months later.

If you want to see the difference on a single page right now, the free Raw vs Rendered HTML checker does a one-off version of the same comparison.

Issues, Not Findings

Site Health does not hand you a list of every imperfection on every page. It opens an issue when a problem spreads across pages, and closes it again by itself once those pages pass.

There are more than 80 issue types, grouped the way you would think about them:

Indexability – noindex added, canonicals pointing to another page or another domain, robots.txt blocking Google.
Errors and redirects – server errors, pages not found, redirects to the homepage, redirect loops and long chains, not-found pages served as normal pages.
Rendering – lost page content, blank renders, menu and footer links removed, empty product or post lists, broken images.
SEO elements – missing main headings, titles and meta descriptions removed or replaced, structured data removed, broken language links.
Security, Performance, and Links and crawl paths – from pages downgraded to HTTP to a slower server response or a broken menu link.
AI crawler access – robots.txt blocking the AI search bots you want, and snippet limits added.

Each issue has a severity, Critical, High, Medium or Low, and the ones to start with are marked Fix first. Open an issue and you see why it matters, how to fix it, the affected pages, and what happened to their search traffic since it started. When you have shipped the fix, I've fixed it – check again rechecks up to 10 affected pages straight away instead of waiting for the next pass.

Some sites do odd things on purpose. Mark an issue as This is normal for one case or for every page, and it stays quiet.

A Score You Can Explain

The Overview opens on a health score from 0 to 100: 90 and above is Good, 70 to 89 Needs attention, and below 70 Poor. It is built from five areas, reachability, crawlability, indexability, experience, and content and SEO, and the score history chart shows the last 30 days. Hover a day and you see which issues moved it, and by how much.

I was strict about what does not count. Outages belong to Uptime Monitoring, so they are left out. So are issues you marked as normal, and sources you have not connected show as "Not measured" rather than as a penalty.

Ranked by What It Costs

Most audit tools sort by severity. Severity says how bad a problem is in general. It does not say how much it costs you.

So every issue in Site Health shows the Google clicks its pages earned in the last 90 days, from Search Console. In the Pages table, every page also shows its last Googlebot visit, from Server Log Monitoring, and whether it is in Google's index. A noindex on 18 blog posts that earn 1,240 clicks is a different afternoon from a noindex on 18 tag pages that earn none, and now the difference is on the screen.

The issue list also pulls in what your other monitors found, from poor page experience on phones to pages missing from Google or an expiring SSL certificate, so there is one list for the whole site. And any page in Site Health can be added to Page Monitoring in one click, if it deserves a closer watch.

Can AI Crawlers Get In?

A separate AI readability page answers a question I get asked a lot now: can AI search actually read my site?

It checks 18 AI and search bots, from OAI-SearchBot, Claude-SearchBot and PerplexityBot to GPTBot and Bingbot, and for each one it shows what your robots.txt allows and, with Server Log Monitoring connected, what your server actually does. They often disagree. A firewall rule or bot protection can refuse a bot your robots.txt invites in. For every AI search or assistant bot that is shut out, there is a copyable fix. Further down, it shows which sections of your site read fine without JavaScript and which depend on it.

Verified Sites Only

Site Health only runs on a domain you have verified as yours, through Google Search Console, Cloudflare, a DNS record, or a meta tag or file. We never crawl a site for someone who has not proved they run it.

Our crawler names itself in its User-Agent, and the bot page explains what it is and how to allow or block it. If your firewall refuses it, the Crawler access settings tell you, and walk you through allowlisting it on the common firewalls and bot-protection services.

Issue history is kept 12 months on paid plans and 14 days on Free.

The Circle Is Closed

This is the third launch in three days. On Sunday it was Sitemap Monitoring, which watches every URL that enters or leaves your sitemaps. Yesterday it was Page Monitoring, which watches the pages you care about most for every SEO change. Today it is Site Health, which watches everything else.

With them, VitalSentinel covers everything I believe you need to monitor a website, and I mean that literally. It started as my answer to running five tools for one site. It is now 12 modules, ten monitors and two analytics:

Core Web Vitals from real users with RUM Monitoring, plus CrUX and Synthetic Monitoring.
Indexing, robots.txt and sitemaps, with Server Log Monitoring showing what the crawlers really did.
Site Health across the whole site and Page Monitoring on the pages that matter most.
Uptime and SSL.
Web Analytics and Google Search Console for the traffic all of it protects.

Every module is on every plan, including Free. The limits are about how much, never whether. The pricing page has the numbers.

Where to Start

Open a verified domain and choose Site Health Monitoring in the sidebar. The Overview tells you how many pages it has checked so far and when your first score should be ready, and from then on it keeps going.

The Site Health introduction is the best place to read first, followed by the full list of issue types and the AI readability guide. The feature page has the overview, and the alert types list what you can be told about.

When it opens its first issue on your site, I would love to hear what it found.

Introducing Site Health Monitoring: A Technical SEO Crawl That Never Stops