What a Technical SEO Audit Should Actually Check
A technical SEO audit becomes noisy fast when teams try to include everything. The better approach is to focus on the checks that explain whether pages can be crawled, indexed, interpreted correctly, and maintained safely over time.
- Technical SEO
- SEO Audit
- Indexing
- Site Health
What Should a Technical SEO Audit Actually Check?
A real audit checks five layers that change decisions: the response layer, crawlability, indexability, canonical and duplication signals, and page-quality context. Everything else is reference material. If a check cannot move a fix from “maybe” to “do this,” it does not belong in the first pass. Scope is the whole job.
Key Takeaways
- Audit five layers: response, crawlability, indexability, canonical and duplication, page quality.
- Crawlability and indexability are separate. Google says robots.txt “is not a mechanism for keeping a web page out of Google” (Google Search Central, 2025).
- Crawl-budget analysis only matters above 1,000,000 pages, so most sites should cut it (Google Search Central, 2024).
- Cut dead checks like dynamic-rendering and
follow,indexdirectives that change no decision.- Run the Technical SEO Audit first, then route to specialist tools.
Most audit problems are not skill problems. They are scope problems. Teams include every check a tool can run, then drown the real findings in a list nobody reads twice. The fix is to decide, before you start, which signals can actually change what you do next.
Why Audit Scope Matters More Than Depth
Scope beats depth because a deep audit of the wrong things still produces no decisions. The check that surfaces a sitewide template fault is worth more than fifty cosmetic warnings. Across the web, 10.6% of desktop pages ship invalid HTML inside the <head> (Web Almanac, 2024). That single fault can hide your canonical and robots tags from crawlers.
The layers stack in a dependency order, and that order is the point. If the response layer is broken, nothing above it is trustworthy. If the page cannot be crawled, indexing signals never get read. So you check from the bottom up, and you stop adding checks once they stop changing decisions.
A wide audit feels thorough. A scoped audit is useful. Those are not the same thing. The Technical SEO Audit keeps the first pass tight on purpose: one target URL plus a small internal sample, enough to expose a pattern without burying it.
Which Checks Belong in Every Audit?
Five checks earn their place because each one can independently change a fix decision. The response layer, crawlability, indexability, canonical consistency, and page-quality context cover the failures that suppress rankings at scale. Only 43% of mobile sites pass all three Core Web Vitals (Web Almanac, 2024), which is why performance context still earns a spot.
The response layer
Check that the URL resolves cleanly, returns the status code you expect, and is not bouncing through redirects. A 200 that should be a 301, or a 301 that chains four times, both change what you fix. The raw response and any redirect hops are the first thing to confirm before you trust anything above this layer.
Crawlability
Confirm that important routes are reachable and not blocked by robots.txt or buried by a config change. Crawlability answers one question: can a bot get to the page at all? A blocked path that should be open is a fast, decision-changing find.
Indexability
This is a separate layer, and conflating it with crawlability causes real mistakes. A page can be crawlable and still carry a noindex, or be blocked and still get indexed from external links. Noindex appears on just 4.7% of desktop pages (Web Almanac, 2024), so when you find it on a page that should rank, it matters.
Canonical and duplication
Check that the page tells one consistent story about which URL should rank. Canonical tags now cover 69% of desktop pages (Web Almanac, 2024), but coverage is not consistency.
Page-quality context
Pull Core Web Vitals and template-level performance signals to inform priority, not to dominate it. In our experience, performance findings are most useful when they point at a shared template, not a single slow image.
Why Are Crawlability and Indexability Separate Checks?
They are separate because the controls live in different places and fail in different ways. Robots.txt governs access. The noindex directive governs indexing. Google is blunt about this: robots.txt “is not a mechanism for keeping a web page out of Google,” and a blocked page can still be indexed if other sites link to it (Google Search Central, 2025).
The trap is that blocking a page in robots.txt actively prevents the fix. If you want a page out of the index, the crawler must read its noindex. Block the page, and the crawler never sees that tag, so the page can linger in results indefinitely. An audit that treats these as one layer will recommend the exact move that breaks the outcome.
So your audit should report them as two findings, every time. One answers “can this be reached?” The other answers “should this appear in search, and does the page agree?” Keep them apart, and the routing to a fix stays clean.
Which Checks Should You Cut From the Audit?
Cut any check whose result never changes a decision. The clearest example is crawl-budget micro-optimization. Google states it only matters for sites above 1,000,000 unique pages, or medium sites past 10,000 pages with content changing daily (Google Search Central, 2024). For everyone else, it is noise.
Dead and obsolete checks
Drop dynamic-rendering checks. The HTTP vary header that signaled it collapsed from roughly 13% of pages in 2022 to 1-2% in 2024 (Web Almanac, 2024). The technique is effectively retired, so a flag for it just adds lines to the report.
No-op directives
Flagging follow and index meta-robots values is busywork. They are the default behavior, ignored as no-ops by Googlebot (Web Almanac, 2024). A report that lists them as “present” implies action where none exists.
Findings that need a larger pattern first
Single low-severity hints, isolated metadata wording, and cosmetic inconsistencies belong in a backlog, not the headline. They only matter if a sample confirms the issue repeats. The honest test for any check is simple: write down the action it would trigger. If you cannot, cut it from the first pass.
Structuring Audit Findings for Handoff
Structure findings by the layer they live in and the next validating step they need. That keeps the report a routing layer instead of a wall of warnings. For each real issue, record the page or sample affected, the layer, the likely scope, and the tool that confirms it. Roughly 2% of pages have a canonical altered during rendering (Web Almanac, 2024), which is exactly the kind of conflict a structured check catches.
A clean handoff maps each finding to one next action:
- redirect instability, route to the Redirect Checker
- sitemap coverage gaps, route to the XML Sitemap Validator
- suspicious response or header behavior, route to the HTTP Header Checker
Shape the output for the reader
Developers need findings that map to implementation. Stakeholders need findings that map to business risk. Same data, different framing. A report shaped for nobody in particular gets read by nobody in particular.
Note the sample scope on every finding
If the Technical SEO Audit flags the same canonical or redirect pattern across several sampled pages, mark it as systemic. One page is a fix. A pattern is a template change, and that distinction decides who owns the work.
What About AI Crawlers and Newer Signals?
AI-crawler directives like GPTBot and ClaudeBot are appearing in more robots.txt files, but they sit outside a classic indexability audit. They govern content licensing and access for AI systems, not whether Google ranks your pages. Treat them as a separate review so they do not blur the search-indexability picture your audit is meant to clarify.
The same discipline applies to every shiny new metric. Before you add a check, ask which decision it changes. INP replaced FID as a Core Web Vital in March 2024, and most mobile sites already pass it while LCP lags behind. That tells you where to spend attention: the metric that fails, not the metric that is newest.
FAQ
Does crawl budget belong in a technical SEO audit?
For most sites, no. Google says crawl budget only becomes a real concern above 1,000,000 unique pages, or for medium sites past 10,000 pages changing daily (Google Search Central, 2024). Below that threshold, the analysis rarely changes a decision, so cut it from the first pass.
Are crawlability and indexability the same check?
No. Crawlability asks whether a bot can reach the page. Indexability asks whether the page should appear in search and whether it agrees. Google confirms robots.txt “is not a mechanism for keeping a web page out of Google” (Google Search Central, 2025), so report them separately.
Should Core Web Vitals be part of the audit?
Yes, as context, not the centerpiece. Only 43% of mobile sites pass all three Core Web Vitals (Web Almanac, 2024), so the signal is common enough to matter. Use it to prioritize template work, not to outrank crawl and indexing problems.
Why flag invalid HTML in the head?
Because it can break the signals you depend on. Roughly 10.6% of desktop pages have invalid HTML in the <head> (Web Almanac, 2024), which can truncate the head and stop crawlers from reading canonical, hreflang, and robots tags. That is a decision-changing fault worth its own check.
What makes a check worth cutting?
A check is worth cutting when its result never triggers an action. Flagging default follow,index directives is the cleanest example, since Googlebot treats them as no-ops (Web Almanac, 2024). If you cannot name the fix a check would prompt, it belongs in reference notes, not the audit.
What To Do Next
Run the Technical SEO Audit first to get the five-layer picture on a target URL and small sample. Then route deeper only where a finding earns it: the Redirect Checker for hop problems, the XML Sitemap Validator for coverage gaps, and the HTTP Header Checker for response anomalies. To decide what gets fixed first, read How to Prioritize Technical SEO Issues After an Audit. To run the pass quickly, read How to Run a Technical SEO Audit Without Wasting Time.