Issues

What the audits do

When a crawl finishes, CrawlStat runs 45 audits over everything it collected: 35 SEO checks and 10 accessibility checks. Each audit looks for one specific problem, such as a broken internal link, a duplicate title or an image without alt text, and records every URL it affects.

Findings come in three severities:

Audits are smart about context. For example, CrawlStat works out once whether each page can be indexed, and checks like title length are only applied to pages that can. You won't be told off for the title of a page that says noindex.

The Issues tab

Open a crawl and click its Issues tab, or All N issues under the top issues on its Overview. The header sums up the crawl, for example "3 errors, 12 warnings, 9 notices".

On the left is every issue the crawl has, grouped under Errors, Warnings and Notices. Within each group, issues are ranked by what fixing them is worth (the same impact score as the plan), then by how many pages they touch, so the top of each group is where to start. Each issue shows a bar with the share of pages it affects, for example "12% of pages", and a Template badge when the same fix repeats across its pages: probably one template change.

On the right is the selected issue:

• Its severity, the number of URLs affected, and Template fix when it applies.
• A description of the problem and why it matters.
• Three figures that say why it ranks where it does:
  • Impact, out of 100: what fixing it is worth, against everything else in this crawl.
  • Reach: the share of pages it affects, and the share of internal link value sitting on them.
  • Effort: how many separate changes it takes, in words, such as "One change fixes all 34" or "Each page needs its own fix".
• A How to fix box with practical advice. It's the same text as the issue's page in these docs.
• Three actions:
  • Show all N in Pages opens the Pages table on the Internal tab, filtered to just the affected URLs.
  • Export affected saves the affected URLs to a CSV file.
  • Copy docs link copies a link to the issue's page on crawlstat.com, which is handy for sending to a developer.
• Affected pages: the first 50 pages, best linked first, so the ones that matter most are at the top. Each shows why it was flagged, for example the duplicate it shares a title with or the broken URL it links to. Click one to open it in Pages.

From issue to fix

A typical workflow:

Running the audits again

Click Rerun audits in the header to run every audit again over the open crawl. The button shows "Auditing…" while it works. Running them again doesn't fetch anything from the site. It reads again what's already in the crawl file.

This is useful in two situations:

• You turned off "Run audits when a crawl finishes" in Settings. The Issues tab will be empty until you run them.
• You've updated CrawlStat. The checks are built into the app rather than stored in the crawl file, so running the audits again on an old crawl gives you the newest version of every check.

If a crawl has no issues listed, the tab says No issues found: either the site is in good shape, or the audits haven't run yet. Click Run the audits to make sure.

Exporting issues

• Export CSV in the header exports a summary with one row per issue: its name, severity, number of URLs affected, description, how to fix it, and a link to its docs page. It makes a ready made to do list.
• Export affected on a single issue exports that issue's URLs, along with a detail column (for example, how many hops a redirect chain has).

Files go to your export folder (Downloads by default), just like Pages exports. To send the findings to a client instead, export a client report.

Thresholds

A few audits use thresholds. These are the defaults:

Character limits for titles and descriptions are a guide. Search engines actually truncate by pixel width, which varies with the characters used.

A few notes on specific audits

• Near duplicate content compares pages by their text. Pages with fewer than 200 words are skipped, because the comparison is unreliable on thin content. Exact duplicates of any length are still caught by Duplicate content.
• Orphan pages are pages CrawlStat knows about, usually from your sitemap, that nothing on the site links to. Orphan detection relies on reading the sitemap, so it works best on a full Spider crawl of a site that publishes one.
• Oversized images needs images to be crawled so it knows their real sizes. Image crawling is on by default.
• Pages blocked by robots.txt lists URLs that your own pages link to, but that robots.txt stops CrawlStat (and search engines) from fetching.

Every audit

Each audit has its own page with a full explanation and fix. A few to start with:

• Errors: Server errors (5xx), Client errors (4xx), Broken internal links, Redirect loops, Mixed content
• Warnings: Missing page titles, Duplicate page titles, Missing H1, Images missing alt text
• Accessibility: Links without accessible names, Form fields without labels, Pages without a language