How We Audit Technical SEO

Why Audits Tend to Be Bad

Most technical SEO audits are 60-page documents that read like a tool output report. Every issue is listed, severity is uniformly "high", and there is no clear sense of what to fix first or whether the issues actually matter for the business reading the document.

The format exists because it is easy to produce and looks thorough. The problem is that nothing gets done with it. The client receives the audit, glances at the page count, and files it next to the strategy deck nobody reads.

What We Actually Do

A technical audit for us has three phases. The first is wide and tool-driven; the second is narrow and human; the third is the hardest and the most useful.

Phase 1: Crawl and capture

We run a deep crawl of the site using Screaming Frog, paired with whatever Search Console history is available. The crawl surfaces the obvious technical issues: broken links, missing canonicals, redirect chains, indexability problems, structured data errors, page speed flags, mobile rendering issues, hreflang inconsistencies if relevant. Anything a tool can find, a tool finds.

This phase is fast and not where the value is. The output of this phase is a long list of issues. The list itself is not the audit.

Phase 2: Manual investigation

The crawl tells us where to look. Then we look manually. This is where the audit starts to become useful. We open the actual pages, look at how they render, look at how the markup is structured, look at how the internal linking is laid out, look at how the URL structure has evolved over time. We read content, we click through journeys, we try to use the site the way a customer would.

Manual investigation surfaces things tools cannot see: pages that technically work but are confusing for users, internal links that exist but make no sense, content that is technically indexable but not actually about anything specific, schema markup that validates but does not match what the page is really about. Most of the meaningful findings in an audit come from this phase, not from the crawl.

Phase 3: Prioritisation

The hardest part. By the end of the manual investigation we have a long list of issues, real and tool-flagged. The job is to figure out which ones actually matter for this specific business.

We sort issues into three buckets. Quick wins: high impact, low implementation effort, do this week. Medium-term work: meaningful impact but involves several days of effort or developer time, do this quarter. Strategic items: significant work that will pay back over months but requires a real commitment, plan into the roadmap.

We also identify issues that look bad in the audit but do not matter for this site. A long list of theoretical SEO problems with no commercial impact is not useful. The audit has to make the call about what to ignore as well as what to act on.

What We Look At

The standard checklist, with comments on what tends to matter most:

• Crawlability and indexation: can search engines actually access and store your pages? Often the most consequential category and often overlooked.
• Internal linking: how authority flows around the site, which pages are buried, which are over-linked. Significant impact and usually easy to improve.
• Site architecture and URL structure: whether the site's organisation matches how customers think about your business, and whether URLs reflect that.
• Structured data: what schema is implemented, what is missing, whether it is correct and matches what the page is about.
• Core Web Vitals: LCP, INP, CLS. Real ranking factor, especially on mobile, and usually addressable with focused work.
• Mobile rendering: not just "mobile-friendly" but how the site actually behaves on a phone in a poor signal area.
• Canonical strategy: whether duplicate or similar content is handled cleanly. Easy to get wrong on dynamic sites.
• Sitemap and robots.txt: often misconfigured, sometimes excluding pages the site owner did not realise.
• JavaScript dependencies: whether content is rendered server-side or client-side, and whether search engines can see it.

What We Ignore

Just as important as what we look at. Some things the SEO industry obsesses over are not worth your attention:

• Domain authority scores: Moz and Ahrefs both publish their own metrics. Neither is a Google ranking factor. Useful for rough competitive comparison, useless as a target to optimise for.
• Keyword density: stopped mattering 15 years ago.
• Meta keyword tags: ignored by Google since 2009.
• Theoretical issues with no traffic impact: a deeply nested page with no commercial relevance is not a priority just because the crawler flagged it.
• PageSpeed scores in isolation: the score is a directional indicator. The actual Core Web Vitals field data is what Google uses.

What the Deliverable Looks Like

A written document, typically 30 to 60 pages depending on site complexity. Structured so technical recommendations can be handed straight to a developer and content-related recommendations straight to a writer or content team. Ranked by impact, not by tool category. Includes a one-page executive summary at the front so a non-technical stakeholder can see the headline findings without reading the rest.

Plus a walkthrough call (or Loom video, depending on preference) talking through the findings and answering questions before any work starts. This matters because the document on its own is harder to act on than the document plus a conversation about what to do first.

Why This Matters

Technical SEO is the foundation. Content and link investment quietly leak value when the underlying technical foundation is broken. A good audit is the cheapest way to figure out what is leaking and what to fix first. The reason most audits are bad is that producing one that genuinely helps a business is harder than producing one that looks thorough.