Client reports
[object Object]
What it is
A client report is a crawl written up for someone who will never open CrawlStat: a client, a manager, a developer on another team. It's one file, either an A4 PDF or a single web page that holds everything, with your name on the cover and your colour throughout.
Exporting one
Open the crawl, then either:
- Click the report button in the crawl's header (its tooltip reads Export a client report), or
- On the Plan tab, click Export PDF under the written report.
Both open the Export report dialog. A small sketch of the cover updates as you fill it in.
| Option | What it does |
|---|---|
| Prepared by | Your name or agency, printed on the cover. Optional. |
| Brand colour | The accent used for headings, figures and charts. Pick a swatch (Indigo, Ember, Ocean, Teal, Forest, Rose or Graphite) or type any hex colour, such as #0F9F8F. |
| Format | PDF: A4, ready to attach or print. Web page: one file that holds everything, opens in any browser and prints to A4 from there. |
| Include the written report | Only offered once you've written one on the Plan tab. It goes in as a summary page straight after the cover. |
| Include progress since the last crawl | Only offered when you've crawled the same site before. On by default. |
CrawlStat remembers your name and colour for next time. Click Export and the file is saved to your export folder (Downloads by default), named after the site, for example report-example-com.pdf, and opened straight away.
What's in it
The report reads top to bottom, the same way you'd explain the crawl in a meeting. Sections are numbered in order; the ones you leave out simply aren't there.
Cover
The site, the date of the crawl, a health score (the share of crawled URLs that answered with a success or a redirect) with a short verdict, and four figures: URLs crawled, Broken or failed, Pages with issues and Link value at risk (the share of internal link value sitting on pages with errors). If the crawl rendered JavaScript, the cover says so.
Summary
If you included the written report, it comes next, exactly as the local model wrote it on the Plan tab. See Local AI.
Overview: where things stand
How the URLs answered, how many kinds of error, warning and notice were found, how many pages can appear in search results (and why the rest can't), the median response time, and how many clicks pages sit from the home page.
Action plan: what to fix first
The top of the plan, in order. Each job shows its impact score, how many pages it touches, how much work it is (one template change, a batch, or page by page), how to fix it, and a few example URLs.
Progress: since the last crawl
When you've crawled the site before: how many issue instances were fixed, how many are new, and which issues got better or worse. It uses the same rules as Comparing crawls, so a page that simply disappeared never counts as fixed.
Accessibility: against WCAG 2.1 AA
The accessibility check's verdict, then the biggest barriers: what each one is, how many pages it affects, and how to fix it.
Appendix: every issue found
Every kind of issue in the crawl, worst first, with how many pages it affects. The full lists of URLs stay in CrawlStat; use Export affected on the Issues tab when a developer needs them.
The report closes with a short note on how it was made: that every figure comes from one crawl, that impact scores aren't a traffic forecast, and that automated accessibility checks don't replace a person.
Tips
- Recrawl before you send. The report describes the crawl you have open. If you've fixed things since, use Recrawl on the Projects screen first, and the report gains a progress section.
- Write the summary yourself if you prefer. The written report is optional. Without it, the report starts at the overview.
- Need the raw data too? CSV exports from the Pages and Issues tabs sit happily alongside a report.