Most “technical SEO audit” guides are tool tutorials in disguise: run the crawler, screenshot the dashboard, done. That produces a list of 300 issues and no idea which three matter. A real audit is a diagnostic process that ends in a ranked action plan – and it hinges on one distinction most guides skip entirely.
This is the professional, paid-tool framework. If you’re a small business owner wanting the free-tool version, start with the DIY technical SEO audit for small business on GrowWithSakib instead. And for the concepts underneath it all, see what technical SEO actually is on GrowWithSakib. This article turns the full technical SEO strategies guide on GrowWithSakib into an auditable checklist.
Before You Start: The Two Ideas That Separate Pros From Tools
The Toolkit
You don’t need all of these, but a professional audit usually combines three types of tool. Here’s the honest landscape with current pricing:
| Tool | Type | What It’s For | Cost (2026) |
|---|---|---|---|
| Google Search Console | Field data | The only source of real Core Web Vitals and index data Google actually uses | Free |
| Screaming Frog | Desktop crawler | The industry-standard deep crawl; the raw ground truth | Free to 500 URLs; GBP 199/yr |
| Semrush Site Audit | Cloud crawler | Automated checks, health scoring, scheduled monitoring | Paid suite |
| Ahrefs Site Audit | Cloud crawler | Technical health plus best-in-class backlink data | Paid suite |
| Sitebulb | Desktop crawler | Visual architecture maps and prioritised hints | Paid |
| PageSpeed Insights | Performance | Core Web Vitals lab + field data, per page | Free |
The 10-Area Audit Framework
Work these in order – the early steps produce the data the later ones depend on.
Area 1: Crawlability – Can Bots Reach Your Pages?
Configure your crawler with Googlebot Smartphone as the user agent (mobile-first indexing means that’s what matters), then crawl the whole site. Cross-reference with Search Console’s crawl stats.
- Look for: blocked resources, dead ends, pages buried too many clicks deep, and crawl-rate problems on slow servers.
- Tool: Screaming Frog for the crawl; GSC Crawl Stats for what Googlebot actually does.
Area 2: Rendering – Can Google See the Content?
The step that saves careers. For a sample of your important templates, run URL Inspection in Search Console and view the rendered HTML. Does the actual content appear, or an empty frame? For JavaScript-heavy sites, configure Screaming Frog with JS rendering and compare the raw vs rendered DOM.
Area 3: Indexation – The Right Pages, and Only the Right Pages
Cross-reference your crawl against GSC’s Pages (index coverage) report. You’re hunting two opposite errors: important pages excluded from the index, and junk pages included (filters, parameters, staging URLs). Both waste crawl budget and dilute signals.
Area 4: Core Web Vitals – Real-World Performance
Use GSC’s Core Web Vitals report (real field data) as the source of truth, then diagnose individual pages with PageSpeed Insights. The three current metrics:
- LCP (loading) – under 2.5 seconds.
- INP (responsiveness) – under 200 milliseconds. This replaced FID in March 2024 and is the most commonly failed of the three.
- CLS (visual stability) – under 0.1.
Area 5: HTTPS and Security
HTTPS has been a ranking signal since 2014 and is now a baseline expectation, not a differentiator. Crawl for any HTTP resources or internal links, mixed-content warnings, and redirect issues on the HTTP-to-HTTPS path. Check GSC’s Security Issues report.
Area 6: XML Sitemaps
Your sitemap is an invitation, not an instruction. It should contain only canonical, indexable (200-status) URLs – no noindex, redirected, or blocked pages. Split files over 50,000 URLs, and keep lastmod dates honest. Confirm it’s submitted in Search Console.
Area 7: Robots.txt
The highest-risk file on your site: one wrong line can hide everything. Confirm you’re not blocking important content or rendering resources (CSS/JS Google needs to render the page). The classic disaster is a “Disallow: /” left over from staging.
Area 8: Canonical Tags
Canonicals resolve duplicate content by naming the master version. Look for conflicting canonicals, chains, pages canonicalising to redirects or noindex pages, and parameter duplicates. Screaming Frog flags conflicts other tools miss.
Area 9: Internal Links and Site Architecture
Run the link report to find orphan pages (no internal links pointing to them), pages buried too deep, broken internal links, and redirect chains. Internal links are how authority flows through your site – and how Google discovers your pages in the first place.
Area 10: Schema and AI Readiness
Validate structured data with Google’s Rich Results Test, and check the right schema types exist on the right templates. Then the 2026 addition: AI readiness. Because AI crawlers like GPTBot and PerplexityBot don’t execute JavaScript, confirm your core content exists in the raw HTML. If it only appears after JS runs, you’re invisible to AI answer engines – detail in the generative engine optimisation guide on GrowWithSakib.
Turning Findings Into a Plan: The Impact x Effort Triage
This is where an audit becomes valuable. Don’t hand anyone a 300-row spreadsheet. Score each issue type (not each URL) on two axes – business impact and fix effort – and sort into three tiers:
| Tier / What Goes Here | Examples | When |
|---|---|---|
| Tier 1 – High impact, any effort – especially high-impact/low-effort | Robots.txt blocking a category; noindex on money pages; rendering failures | This week |
| Tier 2 – Medium impact, reasonable effort | Redirect chains; missing schema; slow key templates; orphan money pages | This month/quarter |
| Tier 3 – Low impact or very high effort | Minor CWV tweaks on low-traffic pages; cosmetic canonical cleanups | Backlog |
The Pro Layer: Log-File Analysis
The one thing no site-audit tool can tell you: what Googlebot actually does on your site. Site-audit tools show you what could be crawled; your server logs show what was. Analysing the last 30 days of logs (with Screaming Frog’s Log File Analyser or a BigQuery pipeline) reveals:
- Crawl-budget waste – Googlebot burning time on parameters, filters, or dead sections.
- Neglected pages – important URLs Googlebot rarely or never visits.
- Silent errors – status codes Googlebot hits that no crawler surfaced.
- Response-time issues – sections Googlebot slowed down on, or abandoned.
How Often Should You Audit?
| Cadence | What to Do | Why |
|---|---|---|
| Monthly | Check GSC coverage, Core Web Vitals, crawl stats; watch for regressions | Plugins, publishing and small changes break things quietly |
| Quarterly | Full 10-area audit with a fresh crawl | Catches drift a monthly glance misses |
| On every major change | Audit within 7 days of a migration, redesign, or CMS change | Migrations routinely lose 30-60% of traffic when unchecked |
Common Technical SEO Audit Mistakes
| Mistake | Why It Hurts | Do This Instead |
|---|---|---|
| Treating the crawl as the deliverable | 300 issues, no priorities, nothing gets fixed | End with a triaged action list |
| Checking crawlability but not rendering | A clean crawl can hide empty JS shells | Verify rendering with URL Inspection |
| Trusting the health score | A 92/100 site can lose half its traffic | Verify key findings manually |
| Auditing against FID | It was retired in March 2024 | Audit LCP, INP and CLS |
| Ignoring server logs | You miss what Googlebot actually does | Analyse 30 days of log files |
| Auditing with a desktop user agent | Misses mobile-first rendering issues | Crawl as Googlebot Smartphone |
| Skipping the post-migration audit | Migrations lose 30-60% of traffic unchecked | Audit within 7 days of any big change |
| Fixing low-impact issues first | Effort spent where it won’t move rankings | Sort by impact, then effort |
Frequently Asked Questions
How do I do a technical SEO audit step by step?
Crawl the site with Screaming Frog using the Googlebot Smartphone user agent, then pull Google Search Console data. Next – and this is the step most guides skip – verify rendering, not just crawling, because a page can be reachable while serving Google an empty JavaScript shell. Then work through indexation, Core Web Vitals, HTTPS, XML sitemaps, robots.txt, canonicals, internal links and site architecture, broken links, and schema plus AI readiness. Finish by triaging every finding into a prioritised action list. The crawl is the start of the audit, not the end.
What tools do I need for a technical SEO audit?
A professional audit usually combines three types: a crawler, an analysis platform, and a performance measurer. Screaming Frog is the industry-standard crawler (free up to 500 URLs, GBP 199 per year for unlimited). Google Search Console is free and irreplaceable – it’s the only source of the real Core Web Vitals field data Google actually uses. Semrush or Ahrefs Site Audit add automated monitoring and health scoring, and Sitebulb adds visual architecture maps. For sites under 500 pages, Search Console plus free Screaming Frog covers most of it.
What is the difference between crawling and rendering?
Crawling is a bot reaching a URL and reading its initial HTML. Rendering is the bot executing the page’s JavaScript to see the finished result, the way a browser would. The distinction matters enormously because Googlebot renders in a separate, delayed queue – so a page can be crawled successfully yet, if rendering fails, be indexed as an empty shell. Many sites with clean crawl reports are invisible to Google because their content only appears after JavaScript runs. Always verify rendering with Search Console’s URL Inspection tool.
How do I prioritise technical SEO fixes?
Score each issue type – not each individual URL – on two axes: business impact (does it affect crawling, indexing, or ranking for pages that make money?) and fix effort (how much time or coordination it takes). Sort by impact descending, then effort ascending, into three tiers. Tier 1 is high-impact work done this week; Tier 2 is medium-impact work for the month or quarter; Tier 3 is a low-priority backlog. Always fix money pages before the blog archive. The order matters more than the length of the list.
How often should I run a technical SEO audit?
Run a full ten-area audit quarterly, with lighter monthly checks of Search Console coverage, Core Web Vitals, and crawl stats to catch regressions early. Crucially, run an immediate audit within seven days of any major change – a platform migration, URL restructure, redesign, or significant CMS update. Migrations are the highest-risk event in technical SEO: unaudited, they routinely cost sites 30 to 60 percent of their organic traffic, and the longer the delay, the harder the recovery.
Can I do a technical SEO audit for free?
Yes, for sites under about 500 pages. Google Search Console (free) covers indexation, Core Web Vitals field data, mobile usability, and security. Screaming Frog’s free tier crawls up to 500 URLs and finds broken links, redirect chains, missing metadata, and duplicate content. Google PageSpeed Insights (free) covers LCP, INP, and CLS. That combination surfaces most of what matters for a small site. Paid tools earn their keep on larger sites, for automated monitoring, or when managing multiple client sites – not because they find things the basics miss on a small site.
Why did my traffic drop even though my audit tool shows a high score?
Because tool health scores measure what the tool knows how to check, and miss what they don’t. The most common culprit is a JavaScript rendering gap: the tool sees a fully-formed DOM and awards a high score, while Googlebot sees an empty shell. Real cases exist of sites scoring above 90 while losing half their traffic. Other blind spots include crawl-budget waste and a single robots.txt line blocking a whole section. Always verify a high score against Search Console’s URL Inspection and your own server logs.
What does log file analysis reveal that other tools don’t?
Log files show what Googlebot actually did on your site, rather than what a crawler thinks could be crawled. Analysing the last 30 days of server logs reveals crawl-budget waste (Googlebot spending time on parameters or dead sections), important pages Googlebot rarely visits, error status codes it silently encountered, and sections where slow response times caused it to slow down or stop. No site-audit tool can tell you this, because it’s a record of real crawler behaviour, not a simulation. It’s the layer that separates a deep audit from a surface one.
Key Takeaways
- A technical SEO audit is a diagnostic process that ends in a prioritised action list – not a crawl report with 300 issues and no priorities.
- Two ideas separate a pro audit from a tool tutorial: crawling is not rendering, and the crawl report is not the deliverable.
- Verify rendering, not just crawling. A page can be reachable yet serve Google an empty JavaScript shell – a real case scored 92/100 while losing half its traffic.
- Work the 10 areas in order: crawlability, rendering, indexation, Core Web Vitals, HTTPS, sitemaps, robots.txt, canonicals, internal links, and schema plus AI readiness.
- Core Web Vitals are LCP, INP and CLS. If your template still lists FID, it’s out of date – INP replaced it in March 2024.
- Triage every finding on Impact x Effort into Tier 1/2/3. Sort by impact descending, then effort ascending, and fix money pages first.
- Log-file analysis is the pro layer: it shows what Googlebot actually crawls, which no site-audit tool can report.
- Audit quarterly, check key reports monthly, and always audit within 7 days of a migration – unaudited migrations lose 30-60% of traffic.
