How to Run a Technical SEO Audit

Step 1: Start the crawl

Open Crawly and click New Crawl. Enter the root URL of the site. For a standard technical audit, use the following settings:

• Mode: Spider (follows all internal links)
• User agent: Googlebot (sees what Google sees)
• Respect robots.txt: On for client sites (mirrors Googlebot behaviour)
• Concurrency: 10 (default - reduce to 3-5 for sensitive servers)

Hit start and let it run. Watch the Pages tab to confirm URLs are coming in. When the crawl completes, work through each section below. Before you start, make sure you have configured your crawl correctly for the site you are auditing.

Step 2: Issues tab

The Issues dashboard is your starting point. Every detected problem is grouped by category and sorted by count. Work top to bottom: errors first, then warnings, then opportunities.

Errors to fix immediately: 4xx broken pages, 5xx server errors, missing title tags, duplicate titles, mixed content. These have the most direct impact on rankings and user experience.

Warnings to address: Missing meta descriptions, missing H1, duplicate H1, redirect chains, non-indexable pages, images missing alt text, missing HSTS, near-duplicate content, hreflang errors.

Click any issue to see the full URL list. Export the list for developer handoff or client reporting.

Step 3: Response Codes tab

Review the status code distribution. A healthy site should be mostly 2xx with minimal 3xx and no 4xx or 5xx.

• High 3xx count - investigate redirect chains (Issues tab will flag these)
• Any 4xx - cross-reference with XML sitemap and backlinks. Which broken pages matter?
• Any 5xx - server errors suggest infrastructure problems or rate limiting. Investigate and retest at lower concurrency.

Step 4: Redirect chains

Back in the Issues tab, expand Redirect chains. Each chain listed shows the full hop sequence. Redirect chains waste crawl budget and dilute link equity - collapse each one to a single direct 301.

Use the Redirect Checker to verify individual URLs are resolving in a single hop after fixing.

Step 5: Image audit

Open the Images tab and filter to Missing alt. This shows every image on the site without an alt attribute.

Alt text matters for both accessibility (screen readers) and image search. The stats bar shows the total count. Export the list for the developer or content team - include the source page URL so they know which pages to update.

Step 6: Security headers

Open the Issues tab and look for security header issues: missing HSTS, missing Content-Security-Policy, missing X-Frame-Options, missing Referrer-Policy, missing X-Content-Type-Options.

These are not direct ranking factors, but missing HSTS is a common finding and a straightforward server configuration fix. Including security headers in your audit report demonstrates thoroughness and gives the technical team quick wins.

Step 7: hreflang audit

For international sites, check the hreflang issues section. Crawly checks for missing x-default, malformed language codes, and hreflang tags pointing to non-200 URLs.

If the site is single-language and has no hreflang at all, skip this section.

Step 8: Site Structure tab

Open the Site Structure tab to see the site as an expandable folder tree. Check that your most important pages (product pages, service pages, landing pages) are close to the root - within 2-3 clicks of the homepage.

Pages buried at depth 6 or deeper get less crawl budget and fewer internal links. If key pages are deep, recommend adding internal links from higher-level pages.

Step 9: Export and report

From each relevant tab or issue, export the URL list. Prioritise findings by impact:

1. 4xx and 5xx errors - immediate fix
2. Non-indexable pages that should be indexable - high priority
3. Redirect chains - medium priority
4. Missing titles and H1s - medium priority
5. Missing alt text - medium priority (also an accessibility issue)
6. Missing security headers - low-medium (quick server config fix)
7. Title and meta description length issues - ongoing content work

Accelerate with Claude Code

If you have Crawly's MCP integration connected to Claude Code, you can skip the manual tab-by-tab triage for the initial analysis:

Give me a summary of the top technical SEO issues on this crawl, prioritised by severity.

Claude calls get_issues and get_summary and returns a structured breakdown in seconds. You still need to review the details, but the initial triage is instant.