Free tool
Check what your website is communicating
One address, one minute, and a genuine assessment of how search engines, generative systems and first-time readers are likely to be interpreting your site.
Passive analysis only. Sensitive technical findings require domain ownership verification.
What this is
A snapshot, not an audit — and the difference matters
This runs the same public analysis that sits behind every IXSEO engagement. It fetches your home page and a few pages linked from it, reads the configuration your domain publishes, measures performance through Google’s own data where available, and then interprets what it found against what the site appears to be trying to achieve.
What comes back is a prioritised set of findings, each written the same way: what was found, why it matters commercially, and the general direction to take. That last part stays general on purpose. Precise instructions require knowing things a public check cannot see — your platform, your constraints, what was tried before and why it was abandoned.
An audit adds exactly that. A consultant reads the same evidence alongside your commercial priorities and your own analytics, then decides which three of the findings would actually change anything. The snapshot is a competent starting position. It is not a substitute for judgement, and it does not pretend to be.
Coverage
What the snapshot assesses
Four modules run by default. The rest are optional, because a business with no email infrastructure concern should not have to sit through a DNS assessment to see whether its titles are working.
Search visibility
- Whether pages can be discovered, crawled and indexed
- Title and description presence, length and usefulness
- Canonical declarations and whether they are self-consistent
- Heading structure and whether it reflects the page's argument
- Structured data presence and declared types
- Sitemap and robots.txt availability and correctness
Generative readiness
- Whether the organisation is defined as a named entity
- Whether services are stated rather than implied
- External corroboration through declared profiles
- Author and expertise signals
- Definitional content a system could quote
- Presence of an llms.txt declaration
Content
- Depth across the pages examined
- Whether headings describe what the page contains
- Thin or near-duplicate page indicators
- Readability of the sampled prose
- Whether content states anything specific about the business
Technical health
- HTTPS enforcement and redirect behaviour
- Measured performance on mobile and desktop where available
- Response time and page weight
- Viewport and language declarations
- Redirect chain length
Optional modules
- Brand consistency and trust presentation
- Competitor comparison against up to three domains
- Email authentication posture
- Public DNS and infrastructure configuration
- Passive security response headers
- Public exposure indicators
What is never done
- Authentication attempts of any kind
- Port scanning or service enumeration
- Vulnerability probing or exploitation
- Form submission or state-changing requests
- Crawling beyond a small sample of linked pages
Sequence
What happens after you press analyse
Eight stages, reported honestly as they complete. The progress you see is real; nothing is padded to make the wait feel productive.
The address is validated twice
Once in your browser for immediate feedback, then again on the server against DNS. The second check is what actually prevents a request being made to a private or internal address, and it is the one that counts.
Public pages are fetched
The home page first, then robots.txt, the sitemap, and a small number of internally linked pages. Every request runs through a guarded fetcher with a byte cap, a timeout and a redirect limit.
Deterministic signals are extracted
Markup is parsed passively — no scripts execute, nothing is re-rendered. Facts such as whether a title exists, how many H1s there are, and which structured data types are declared are established by code, not inferred.
External checks run in parallel
Performance data, DNS records and header analysis are requested concurrently. Any of them can fail without stopping the analysis; the report says which checks did not complete rather than quietly omitting them.
Evidence is assembled
Everything gathered is collected into one structured object. This is the only thing that goes forward — no raw page content is carried into interpretation.
Findings are interpreted
Rule-based interpretation always runs and establishes the factual findings. Where a language model is configured it adds the interpretive layer — priority, business framing, executive summary — and its output is validated against a strict schema before anything is displayed.
The public snapshot is filtered
Findings that would expose sensitive technical detail about a domain you have not verified are withheld, with their presence stated rather than hidden. What remains is what any member of the public could establish for themselves.
Results are shown
Scores with explanations, the strengths and weaknesses that stood out, the most urgent issue, and a selection of findings in full.
IXSEO performs passive public analysis unless a separate authorised engagement has been agreed.
Responsible analysis
The limits, stated plainly
This analysis reads a static document. If your site renders its content through JavaScript after load, much of what a visitor sees will be invisible here. That is worth knowing in itself, because some crawlers have the same difficulty, but it means the content findings may understate what is actually published.
It sees a handful of pages, not your site. Conclusions about coverage and depth are drawn from a sample, and the report says so. A large site with excellent section pages and a sparse home page will read worse here than it deserves.
It has no access to your own data. Search Console and analytics show what people actually did, which is the only way to separate a page nobody finds from a page people find and immediately leave. Without them, some conclusions are necessarily inferential, and those are marked with lower confidence.
And it cannot prove an absence. A perfect score is not available here, because passive analysis establishes what is present, not that nothing is wrong. Anything scoring in the nineties has no observable problems in the things we could check — which is a genuinely good position, and a different claim.
Common questions
Is this the same as the homepage analysis?
How many pages does it look at?
Why does my score not match another tool's score?
Will running this put us on a mailing list?
The analysis failed on our site. What went wrong?
Narrower checks
If you already know which question you have, these answer it directly.
Ready for the version a person has read?
A reviewed audit takes the same evidence and decides what to do about it, in what order, against what your business actually needs.
Scores are directional assessments based on the available evidence and should not be interpreted as search-engine rankings.