Fetch
The raw response, as it arrives. Fast, and on a static page it tells you everything you need.
On a static HTML site this is the faster route to the same answer.
Meta Snapshot
What you see is not what the machine receives.
A page can look finished in a browser and arrive at a crawler as something quite different. Meta Snapshot captures what is actually there to be parsed at a given URL, so the conversation moves from what the page should say to what it demonstrably does.
No card required.
- Fetch and render modes
- Page-level evidence
- Comparable across runs
Meta Snapshot Page-level capture
Two ways of looking
Fetch and render can disagree, and the disagreement is the finding.
If a page builds its content with JavaScript, what arrives in the raw response and what exists after rendering are two different documents. Checking only one of them is how teams end up confidently wrong.
Render
The page after it has been built in a browser context, including content that only exists once scripts have run.
On a JavaScript-driven site, fetch alone can make real content look missing.
Picking the wrong mode for the site is the most common way this evidence gets misread.
What gets captured
The signals that decide whether a page can be used at all.
These are not style preferences. Each one can quietly remove a page from consideration, or change how it is understood.
-
Title
How the page identifies itself before anything else is read.
-
Canonical
Which URL the page claims is authoritative — a frequent source of accidental self-exclusion.
-
Robots
Directives that can remove a page from eligibility entirely.
-
Headings
The structure that carries what the page is actually about.
-
Schema
Machine-readable markup describing the entity, product or article.
-
Core HTML
The substance available to be extracted, rather than assembled later.
Before and after
A snapshot is most useful in pairs.
Taken once, it tells you the current state. Taken before a deployment and again afterwards, it tells you whether the change did what was intended and whether anything else moved with it. Template changes in particular tend to alter pages nobody was thinking about.
Capture a baseline before shipping a technical change, not after something looks wrong.
Keeping this in proportion
Metadata does not decide whether AI systems reference you.
It decides whether a page can be reached, parsed and understood — which is a precondition, not a cause. Pages can be technically immaculate and still go unreferenced because they do not answer anything anyone is asking. Treat this as removing obstacles, not as a lever on visibility.
What Meta Snapshot connects to
Page-level capture is the raw material other views summarise or act on.
-
Aggregates it
Technical Audit
The same class of findings, rolled up by severity.
-
Summarises it
AI Health
Readiness across access, indexability and extractability.
-
Acts on it
Fix Assist
A finding turned into implementation-ready work.
-
Compares it
Draft Preview
A draft URL checked against the live page.
See what your pages actually hand over.
Capture a URL and compare what fetch and render return.
Not sure which mode your site needs? Talk to us