Segments
A clean shop and a neglected blog do not average out
Score /blog, /shop and help.example.com separately, from the same crawl and for no extra credits.
- extra credits per segment
- 0extra credits per segment
- or subdomain
- Pathor subdomain
- not shares of one
- Real scoresnot shares of one
- the same slice, crawl to crawl
- Over timethe same slice, crawl to crawl
Why it is one crawl and not several
The obvious way to score a subfolder is to crawl it on its own. That costs credits per segment and fetches your pages more than once - and worse, it cuts the link graph at the segment boundary.
A link from your blog to a product page would be reported as broken, and the product page as an orphan, because the crawl that found the link never saw where it went. Both findings would be false, both would be at the top of the list, and nothing about them would be your site’s fault.
What a segment gives you
- Its own health score - and its own history, so you can watch one part improve while another does not
- Its own ranked fixes - the top problems in that part, which are rarely the site’s top problems
- Its own counts - pages, criticals, warnings, per category
- A section in the client report - so a client sees the part they care about rather than an average
How people slice a site
Blog against the rest
The most common one, and usually the most uncomfortable: the content everyone writes and nobody audits.
Shop, category, product
Templated pages fail in templated ways - one segment tells you whether it is the template or the content.
Documentation on a subdomain
Matched by host rather than path, and often the healthiest part of a site nobody thought to check.
A migration in progress
The new section against the old one, from the same crawl, while both are live.
The same slice, crawl after crawl
A segment lives on the site rather than on a scan, and its score is written by the same code path that writes the crawl’s own - so a rescore moves both together and the segment numbers can never drift away from the score printed beside them.
Questions
- No. A segment is a filter over the crawl you already ran, so ten segments cost exactly what the crawl cost. Crawling each part separately would spend credits per segment, fetch your pages several times, and break the link graph at every boundary - a link from /blog to /shop would read as broken.
- No - it is a real score, produced by the same function applied to that segment's pages. That works because the health score is a per-page rate with per-category caps rather than a sum. It follows that segment scores do not add up to the site score, and should not be read as though they do.
- By path prefix (/blog) or by host (help.example.com). Both are exact and cheap. A regular-expression type is deliberately absent: user-supplied regex over every URL of every crawl is a denial-of-service surface for a convenience nobody has asked for.
- It shows no score rather than 100. Zero pages with zero findings is arithmetically perfect and an absurd claim, so the absence is stored as an absence.
- Yes - as a section you can include or leave out. An agency reporting on a site with distinct areas usually wants it; a one-page site does not.
Start free
One site and 100 credits - up to 200 pages a scan, about 5 scans, and the whole report. No card. Paid plans add more sites, bigger crawls and the AI rewrites.