What it draws on
- A full crawl
- Screaming Frog SEO Spider, run against the live site.
- Google Search Console
- The Page indexing, Core Web Vitals and Performance reports, and URL Inspection. Optional: without access the audit runs on the crawl and lab tests, and says what it couldn’t see.
- PageSpeed Insights
- Field data from the Chrome UX Report for the previous 28 days where Google has enough of it, and a Lighthouse lab test of each key template.
- Rich Results Test
- Google’s own test of which rich results a page’s markup is eligible for.
- axe
- Automated accessibility checks, followed by a manual keyboard pass.
- By hand
- robots.txt, the page source, and a set of questions put to AI search.
CheckWhat we look forChecked with
01Status codes
Every URL the crawl reaches, sorted into 200s, redirects, 404s and server errors, with the pages that link to each broken one.
Crawl
02robots.txt and noindex
What is blocked from crawling and what is kept out of the index, and whether either catches a page meant to rank. A page blocked in robots.txt can still appear in results without a description, so the two are checked together.
Crawl, by hand
03Canonicals and duplicates
Pages with no canonical, a canonical pointing at another URL, and the same content reachable at several addresses. A canonical is a strong signal to Google, and Google can still choose a different URL.
Crawl
04XML sitemap
The sitemap compared with the crawl: URLs listed that redirect or 404, and indexable pages left out.
Crawl
05Redirects
Chains, loops and temporary redirects doing a permanent job. Google advises redirecting straight to the final URL.
Crawl
06What Google has indexed
The Page indexing report’s reasons, read against the crawl: pages crawled and not indexed, soft 404s, pages Google treats as duplicates.
Search Console
07Depth and orphan pages
How many clicks each page is from the home page, and pages no internal link points at. Google asks for at least one link to every page you care about.
Crawl
08Internal links and anchor text
Links that aren’t crawlable anchors, and anchor text such as “click here” that says nothing about the target.
Crawl
09Titles and meta descriptions
Missing, duplicated and near-empty ones, listed by URL.
Crawl
10Headings, images and thin pages
Pages with no H1 or several, headings out of order, images without alt text, and pages with almost nothing on them.
Crawl
11Structured data
What markup is present, whether it validates, whether it describes what is visible on the page, and which types Google supports that the site doesn’t carry yet.
Rich Results Test
12Core Web Vitals
LCP, INP and CLS at the 75th percentile of real mobile visits, against Google’s thresholds: 2.5 seconds, 200 milliseconds and 0.1. Where a site has too little field data, a lab test stands in, and INP can’t be measured in a lab, so the report says so.
PageSpeed Insights
13Content gaps
Searches the site already appears for with no page that answers them, and pages competing with each other for the same search. Needs Search Console access.
Search Console
14Current rankings
Where the site ranks today for up to 20 keywords agreed with you, recorded with the date. It’s the baseline for measuring later work.
By hand
15AI answers
Whether robots.txt lets AI search crawlers in, whether key pages are indexed and open to snippets, and whether the site is cited for a short list of questions agreed with you, in Google’s AI Mode, ChatGPT search and Perplexity.
By hand
16Accessibility
WCAG 2.2 AA failures on the key templates, such as unlabelled form fields, missing alt text and parts that can’t be reached by keyboard.
axe, by hand
17Security
HTTPS on every URL, mixed content and, with a CMS login, core, theme and plugin versions that are out of date.
By hand