Most SEO problems that cost real traffic are not subtle. A template adds noindex to a whole section. A migration leaves thousands of old URLs returning 404. A JavaScript framework ships pages that are empty until scripts run. Each one is easy to fix once you see it, and invisible until you look.
A technical SEO audit is how you look. This guide covers what an audit checks, how to run one step by step, how to decide what to fix first, and how to keep the site from sliding back. Each area links to a detailed guide, and the technical SEO glossary defines the terms.
What is a technical SEO audit?
A technical SEO audit is a systematic check of whether search engines can crawl, render and index a website, and of the technical problems that weaken its pages in search. It’s done by crawling the site the way a search engine does, checking every URL against known issues, and joining the results with data from Google Search Console.
Google’s own baseline is short. Its technical requirements for a page to be eligible for Search are that “Googlebot isn’t blocked”, that “the page works, meaning that Google receives an HTTP 200 (success) status code”, and that “the page has indexable content”. It adds that meeting them doesn’t guarantee indexing.
An audit checks those three things on every URL, then everything that makes pages harder to crawl, slower, duplicated or less useful in the results.
For a separate pass over the baseline itself, use the Google Search Essentials audit: it distinguishes technical eligibility, spam-policy risks and best practices without treating a passed checklist as an indexing guarantee.
If Search Console reports an actual manual action, use the Google manual actions recovery checklist to establish its scope, fix the policy violation across affected pages and prepare a review request. A traffic drop alone is not evidence of one.
What does a technical SEO audit check?
| Area | The question | Detailed guide |
|---|---|---|
| Crawling and indexing | Can search engines reach every important page, and are they allowed to index it? | Crawling and indexing |
| Status codes and links | Do pages return 200? Do links point at working URLs? | Broken links, redirect chains |
| Duplicates and canonicals | Does each piece of content have one clear URL? | Duplicate content, canonical errors |
| On-page elements | Does every page have a unique, descriptive title, description and H1? | Title and meta audit |
| Content quality | Are there pages with too little to offer? | Thin content |
| Site structure | Is every important page linked, and close to the homepage? | Internal linking, orphan pages |
| JavaScript | Is the content there before scripts run? | JavaScript SEO |
| Mobile-first indexing | Does Smartphone Googlebot see equivalent content and page signals? | Mobile-first indexing audit |
| Security | Is every page on HTTPS, with no mixed content? | HTTPS and mixed content |
| Performance | Do pages load fast for real users? | Core Web Vitals audit |
| Structured data | Is markup valid and eligible for rich results? | Structured data errors |
| International | Do language versions point at each other correctly? | Technical international SEO |
| Images | Do images load, and do they have alt text? | Image SEO audit |
| Search performance | Which pages rank but underperform? | Striking distance, organic CTR |
| AI search | Can AI crawlers reach, read and quote your pages? | AI search readiness |
How do you do a technical SEO audit, step by step?
1. Define the scope
Decide what you’re auditing: the whole domain, one subdomain, or one section such as the blog or the shop. Note anything unusual before you start: a recent migration, a new CMS, a section that lost traffic. That tells you where to look first.
2. Crawl the site
Crawl the way a search engine would: start at the homepage and follow internal links. That shows what users and crawlers can actually reach. In Crawlens, the crawl modes cover the three common cases:
| Mode | What it does | Use it for |
|---|---|---|
| Website | Starts at a URL and follows internal links | Most audits |
| Sitemap | Crawls the URLs in your XML sitemaps, optionally also following links | Sitemap quality, finding orphan pages |
| URL list | Crawls exactly the URLs you paste or import | Checking a release or a migration |
Turn on Render JavaScript if the site is built with React, Vue or another client-side framework. Keep Respect robots.txt on, so the crawl sees what search engines are allowed to see.
Already use another desktop crawler? The comparisons with Screaming Frog and Sitebulb show where each tool fits, including where they do more than Crawlens.
3. Add the data a crawl can’t see
A crawl tells you what the site looks like. It doesn’t tell you what Google does with it, or which pages matter. Add:
- Search Console: clicks, impressions and positions per URL, so issues on pages with traffic stand out.
- Analytics: sessions per landing page, from every channel.
- Core Web Vitals: real-user field data and lab scores from PageSpeed Insights.
- Server logs: what Googlebot and other crawlers actually requested.
Crawlens joins all four to the crawled URLs, which turns on checks like Pages with search traffic return errors and Indexable pages Googlebot did not crawl.
4. Read the results by severity
Crawlens runs 70 checks after every crawl, each with a severity:
| Severity | Meaning | Examples |
|---|---|---|
| Critical | Stops pages from being reached or indexed. Fix first. | 4xx and 5xx pages, broken internal links, redirect loops |
| Warning | Hurts crawling, indexing or the search result. Fix soon. | Redirect chains, duplicate titles, invalid structured data |
| Notice | Worth checking; often fine on purpose. | Thin content, deep pages, noindex pages |
The health score sums this up: 100 × (1 − (URLs with a critical issue + 0.5 × URLs with a warning) ÷ all crawled URLs). Notices don’t lower it. It’s useful for tracking progress over time, less so for comparing different sites. The audit rules reference lists every check with why it matters and how to fix it.
5. Prioritise
Severity says how bad a problem is in general. Priority depends on your site. Sort issues by:
- Pages with traffic. An error on a page with 2,000 clicks a month comes before the same error on a page nobody visits.
- Templates. One fix to a template often clears hundreds of URLs. Look for issues that share a URL pattern.
- Effort. Fixing a redirect rule takes minutes; re-platforming a JavaScript site takes months. Do the cheap, high-impact fixes first.
6. Fix and document
Write each fix as a ticket a developer can act on: the issue, the affected URLs (export them), why it matters, and what “fixed” looks like. Crawlens exports an Excel workbook with every issue and its URLs, and a PDF report for clients or stakeholders. See technical SEO audit reports for what to include.
7. Re-crawl and compare
When fixes ship, crawl again with the same settings and use Compare. It shows which issues were fixed, which are new and how the health score moved. Fixes that didn’t work, and regressions the release introduced, show up straight away.
For a smaller pre- and post-release gate, use SEO regression testing to compare search-critical behavior on representative URLs before a change reaches the whole site.
8. Keep it running
Sites change every week. Schedule crawls (Crawlens can crawl automatically, audit, compare with the previous crawl and email you when the health score drops) so new problems surface within days.
The audit, area by area
Crawling and indexing
Start here, because nothing else matters for a page search engines can’t reach or aren’t allowed to index. Check robots.txt rules, noindex directives, canonical tags, status codes and whether every important page is linked and listed in the sitemap.
Key checks: Noindex pages, Linked pages blocked by robots.txt, Canonical points to a non-200 URL, Non-indexable URLs in sitemaps, Orphan pages. The full process is in the crawling and indexing guide, and the most common confusion is explained in noindex vs robots.txt.
Status codes, broken links and redirects
Every internal link should point at a URL that returns 200. Links to 404s and 5xx errors send visitors and crawlers to dead ends; links to redirects add a hop to every request; chains and loops waste more.
Key checks: Internal pages return 4xx/5xx, Pages link to broken internal URLs, Redirect chains (2+ hops), Redirect loops, Internal links point to redirects. See how to find broken links and redirect chains.
Duplicate content and canonicals
When several URLs show the same content, search engines pick one to show, and it may not be the one you want. Parameters, http/https and www variants, trailing slashes and printer versions are the usual causes.
Key checks: Duplicate content, Near-duplicate content, Missing canonical tag, Multiple canonical tags, Canonical points to a noindex page. See how to find duplicate content and canonical tag errors. Online stores have their own sources of duplicates, from filters and variants; see the e-commerce technical SEO guide.
Titles, meta descriptions and headings
The title is the biggest on-page signal you control and the first thing searchers see. Missing, duplicated or truncated titles usually come from templates, which makes them cheap to fix at scale.
Key checks: Missing page title, Duplicate page titles, Title longer than 60 characters, Missing meta description, Missing H1. See the title and meta description audit.
Content quality
Technical audits can’t judge whether content is good, but they can find where it’s missing: indexable pages with almost no text, and pages that never appear in search.
Key checks: Thin content (under 200 words), Indexable pages with no search impressions. See thin content.
Internal linking and site structure
Internal links decide how crawlers discover pages and which ones you signal as important. Pages buried deep or linked from a single place get crawled less.
Key checks: Pages more than 4 clicks from the start, Pages with only one internal link, Pages with no internal links, Internal links with nofollow. See the internal linking audit and orphan pages.
JavaScript rendering
Google renders JavaScript, but later and at a cost, and most AI crawlers don’t render it at all. Compare the raw HTML with the rendered page to see what depends on scripts.
Key checks: Content depends on JavaScript, Links only in rendered HTML, JavaScript changes SEO tags. Start with the JavaScript SEO guide, then use the raw-versus-rendered HTML audit for the comparison.
Mobile-first indexing
Check the mobile render, not only a desktop crawl: primary content, internal links, robots rules, metadata, structured data and images should still be available to Smartphone Googlebot. Compare representative templates with the mobile-first indexing audit, especially when mobile and desktop serve different HTML or URLs.
HTTPS and security
Every page should be served over HTTPS, the HTTP version should redirect, and HTTPS pages shouldn’t load HTTP resources.
Key checks: Pages served over HTTP, HTTP version does not redirect to HTTPS, Mixed content. See HTTPS and mixed content.
Search Console’s Security Issues report is a separate check for detected hacks, malware and harmful downloads. Use the Security Issues audit workflow to verify affected URLs and recovery steps; it is not the same as a manual action.
Page speed and Core Web Vitals
Use real-user data (the Chrome UX Report) where it exists, lab data where it doesn’t, and fix by template.
Key checks: Poor LCP / INP / CLS for real users, Slow server response (over 1 second), Low Lighthouse performance score. See the Core Web Vitals guide and the Core Web Vitals audit.
Structured data
Invalid JSON-LD is ignored entirely; items missing required properties aren’t eligible for rich results.
Key checks: Invalid JSON-LD, Structured data missing required properties. See structured data errors.
International SEO
For multilingual or multi-regional sites, hreflang annotations must use valid codes, point at indexable pages and be reciprocal.
Key checks: Hreflang without return links, Invalid hreflang codes, Hreflang set without x-default. See the technical international SEO guide for the full architecture and hreflang errors for annotation fixes.
Images
Images need alt text to be accessible and understood, and broken image URLs make pages look neglected.
Key checks: Images missing alt text, Broken images. See the image SEO audit.
Search performance
With Search Console joined to the crawl, the audit can show where the quickest wins are: pages just off page one, and pages on page one that searchers skip.
Key checks: Pages ranking just below page one (positions 11–20), Low click-through rate for the ranking position, Pages with search traffic are not indexable. See how to use Google Search Console for SEO, striking distance keywords and how to improve organic CTR.
AI search readiness
AI Overviews, ChatGPT search, Claude and Perplexity can only cite pages they can crawl and read. Check which AI crawlers robots.txt and your CDN allow, and whether your content exists without JavaScript. See the AI search readiness checklist.
When should you run a technical SEO audit?
- Before and after a migration, redesign or CMS change. Crawl the old site to benchmark it, the new one before launch, and both after. See the site migration checklist.
- After releases that touch templates, routing or rendering.
- Before a planned outage or maintenance window: decide whether pages can stay live, verify the temporary response and check recovery. Use the planned website outage SEO checklist for that narrower workflow.
- When traffic drops without an obvious cause.
- On a schedule: a full review quarterly, automated crawls weekly or monthly in between.
How do you go beyond the built-in checks?
Every site has rules of its own: every product page needs a price, every article an author, no page may load an old tracking script. Use custom extraction to pull any value from the HTML into a column, and custom rules to turn it into an issue that shows up in every audit.
You can also let an AI assistant read the audit. Crawlens’s MCP server gives Claude and other assistants access to your crawls, issues and URLs, so you can ask questions like “which templates cause most of the duplicate titles?” See running a technical SEO audit with Claude.
Technical SEO audit checklist
Setup
- Scope defined; recent changes and traffic drops noted
- Full crawl from the homepage, respecting robots.txt
- JavaScript rendering on for client-rendered sites
- Search Console, Analytics, Core Web Vitals and server logs added where available
Crawling and indexing
- robots.txt blocks nothing that should rank
- No unintended noindex
- Canonicals point at indexable 200 pages
- Sitemaps list only indexable 200 URLs
- No orphan pages; important pages within four clicks
Status codes and links
- No internal 4xx or 5xx pages, or links to them
- No redirect chains or loops; internal links point at final URLs
- Pages with traffic all return 200
Content and on-page
- Unique titles and descriptions on every indexable page
- One H1 per page
- No duplicate or near-duplicate pages competing
- Thin pages improved, merged, noindexed or removed
Rendering, security and speed
- Main content and links present in the raw HTML
- HTTPS everywhere; HTTP redirects; no mixed content
- Core Web Vitals pass for real users; server responds in under a second
Markup and international
- Valid structured data with required properties
- Hreflang reciprocal, valid and pointing at indexable pages
- Images load and have alt text
Follow-up
- Issues prioritised by traffic, template and effort
- Fixes written up with affected URLs
- Re-crawl and compare after fixes ship
- Scheduled crawls with alerts
All Technical SEO Audit guides
- Planned website outage SEO: keep a short shutdown safePlan a short maintenance window without accidentally deindexing your site. Choose a limited-function mode or a temporary 503 and verify recovery.
- Google manual actions SEO: diagnose the notice and request reviewCheck Search Console for manual actions, investigate affected URL patterns, fix the underlying policy violation and prepare a review request.
- Google Search Essentials audit: technical checks and policy reviewCheck Google's minimum technical requirements, useful-content basics and spam-policy risks without treating a crawler score as a policy certificate.
- Mobile-first indexing audit: find what Googlebot misses on mobileAudit mobile content, links, robots rules, metadata, structured data and images against desktop pages so Googlebot sees the pages you intend to index.
- Google Search Console Security Issues: Audit and FixHow to investigate the Search Console Security Issues report, find affected pages, fix the underlying problem and request a review safely.
- SEO regression testing: catch search-breaking releasesBuild a release-focused SEO regression test for canonical tags, redirects, robots rules, rendered content and structured data without repeating a full audit.
- Custom extraction for SEO: 8 things to pull from every pageUse CSS selectors and regex to extract prices, authors, dates, analytics tags and schema fields from every page, then flag pages with custom rules.
- How to find broken links on your website (and fix them)Find every broken internal link, external link and 404 page on a site with one crawl, decide which ones matter, and fix them at the source.
- How to find duplicate and near-duplicate content on your siteExact and near-duplicate pages, duplicate titles and descriptions. How to find them across a site and when to redirect, canonicalise or rewrite.
- HTTPS and mixed content: how to find and fix insecure pagesPages still on HTTP, HTTP that doesn't redirect, and images or scripts loaded insecurely on HTTPS pages. How to find them across a site and fix them.
- Image SEO audit: missing alt text and broken imagesHow to find images without alt text and images that fail to load across a whole site, write alt text that helps, and handle decorative images correctly.
- Structured data errors: how to find and fix them across a siteBroken JSON-LD and missing required properties cost rich results. How to audit schema markup across every page, not one URL at a time, and fix it.
- Technical SEO audit report: what to include and how to present itHow to turn a crawl into a technical SEO report people act on: the summary, prioritised issues, evidence, fixes, and how to show progress between audits.
- Thin content in SEO: how to find it and what to do with itWhat thin content is, why word count alone doesn't define it, how to find thin pages across a site, and when to improve, merge, noindex or remove them.
- Title tag and meta description audit: find and fix every problemMissing, duplicate, too long or too short: how to audit titles, meta descriptions and H1s across a site, and rewrite the ones that cost clicks.
Frequently asked questions
What is a technical SEO audit?
A technical SEO audit is a review of how well search engines can crawl, render and index a website, and of the technical issues that hold its pages back: errors, redirects, duplicates, slow pages, broken markup. It's usually done by crawling the whole site and checking every URL against a list of known problems.
How long does a technical SEO audit take?
The crawl itself takes minutes for a small site and hours for a very large one. Reading the results, prioritising and writing up fixes usually takes a day or two for a first audit. Later audits are faster, because you're comparing with a known baseline.
How often should you do a technical SEO audit?
Run a full audit at least quarterly, and a crawl after every release, redesign, migration or CMS change that touches templates or URLs. Scheduled crawls, weekly or monthly, catch regressions between audits.
What is the difference between technical SEO and on-page SEO?
Technical SEO is about whether search engines can access, render and index pages: status codes, robots rules, canonicals, rendering, speed, site structure. On-page SEO is about the content of each page: titles, headings, copy and how well it answers the query. Audits usually cover the overlap, such as missing titles and duplicate content.
What tools do you need for a technical SEO audit?
At minimum, a site crawler and Google Search Console. A crawler checks every URL the way a search engine would; Search Console shows how Google actually indexes and ranks them. Server logs and Core Web Vitals data add evidence for larger sites.