Reading the report that says what is stored

The basic split

Pages are either indexed or not, and the not indexed group is broken down by reason.

The reason is the useful part. A page excluded because you told it to be excluded is a success, not a failure.

So read the reasons rather than the total. A large not indexed number is common and often correct.

The tool also separates pages it learned about from your sitemap from pages it found by following links. A large gap between those two is usually a sign your sitemap is out of date.

A large not-indexed number is common and often correct, so read the reasons rather than the total.

Statuses that are fine

Excluded by a noindex instruction, where you meant it. Pages that redirect elsewhere. Alternate versions with a canonical pointing to the main page.

Not found pages for addresses you deliberately removed. Those should return nothing, and reporting them is the tool doing its job.

Leaving these alone is correct. Chasing them to zero is the most common way to waste a weekend.

• Excluded by a noindex instruction, where you meant it
• Page with a redirect
• Alternate page with a proper canonical tag
• Not found, for an address you deliberately removed
• Blocked by robots, where you intended the block
• Duplicate without a canonical, on a variant you do not care about

Statuses worth investigating

Crawled and currently not indexed means the engine looked and chose not to store it. Usually thin or duplicated content.

Discovered and currently not crawled means it knows the page exists but has not fetched it. Often a sign of a slow site or a large number of low value addresses.

Duplicate with a different canonical chosen means the engine disagreed with your choice. Worth understanding before changing anything.

Soft not found is the subtle one. It means the page returned a normal response while looking empty to the engine, which usually points at a template rendering nothing when content is missing.

Soft not found is the subtle one: a page returning a normal response while looking empty, usually a template rendering nothing when content is missing.

Statuses that are real problems

Server errors. Pages blocked by robots that you intended to be found. Redirect chains that never resolve.

A noindex instruction on pages you want indexed, which is nearly always a leftover setting from staging.

These are the ones to fix today, because they are mechanical faults rather than judgments about quality.

Validation

After fixing something you can ask the tool to check again. It rechecks a sample and reports back over the following days.

Do not start a validation before the fix is actually live, because a failed validation takes longer to retry.

One fix, one validation, then wait. Batching several changes makes the result impossible to read.

One fix, one validation, then wait. Batching several changes makes the result impossible to read.

What a healthy small site looks like

Most real pages indexed, a handful excluded on purpose, and a few in the crawled and not indexed group that you are working on.

There is no target number. A site with three hundred pages and two hundred and eighty indexed is in good shape.

Compare against yourself over time rather than against anybody else.

The shape matters more than the numbers. A steady count with occasional additions is healthy. A count that drops by a third in a week is a fault, whatever the absolute figures are.

How the builder keeps this clean

Every page written by Website Builder Studio has one address, one canonical, a real response for anything missing, and a sitemap rebuilt from what exists.

That removes most of the mechanical entries in this report before they appear.

What remains is the judgment category, which is about whether pages are genuinely useful, and the check that runs before anything publishes measures exactly that.

The sitemap carries a real modification date for each page, taken from the file itself rather than from the moment of the build. That means a date change signals a real content change, which is what the date is for.

One address per page, one canonical, a real response for anything missing, and a sitemap rebuilt from what exists — the mechanics covered in sitemaps.