AD Search Studio

Technical SEO

Technical SEO Audit Checklist: What to Check, and in What Order

Key takeaways

  • Audit in order: crawl, indexation, architecture, internal links, rendering, speed, structured data, then logs.
  • Screaming Frog and Search Console cover most of the audit. Server logs show what Googlebot really does.
  • Core Web Vitals are judged on field data at the 75th percentile: LCP 2.5s, INP 200ms, CLS 0.1.
  • Every finding needs affected URLs, a fix, an owner and a way to verify it, or it will not get done.

A technical SEO audit checks, in order, whether search engines can crawl your site, whether the right pages are indexed, whether your architecture and internal links send authority to the pages that matter, and whether those pages render, load and describe themselves correctly. Run it in that sequence, because a problem early in the chain (a blocked folder, a wrong canonical) makes everything after it irrelevant.

This is the checklist I use across client sites and in my technical SEO work, with the tool for each step and what "good" looks like before I move on.

The audit at a glance

StepMain toolsWhat good looks like
1. CrawlabilityScreaming Frog, Search Console robots.txt and Crawl stats reportsNo important URL blocked, internal links go straight to 200 pages, server errors near zero
2. IndexationSearch Console page indexing report, URL InspectionIndexed pages match the pages you want indexed
3. ArchitectureScreaming Frog crawl depth and Crawl AnalysisMoney pages a few clicks from the homepage, no orphans
4. Internal linksScreaming Frog inlinks and anchor textMost links point at the pages that earn revenue
5. RenderingScreaming Frog JavaScript crawl, URL InspectionContent, links and tags present in the HTML Google renders
6. Core Web VitalsSearch Console, PageSpeed Insights, Chrome DevToolsLCP, INP and CLS all "good" on field data
7. Structured dataRich Results Test, Schema Markup ValidatorValid markup that matches visible content
8. Hreflang (if relevant)Screaming Frog hreflang tabReciprocal, valid codes, pointing at indexable URLs
9. Log filesScreaming Frog Log File AnalyserGooglebot spends its time on indexable pages

Before you start: set up the crawl

Verify the site in Search Console as a Domain property so you see every subdomain and protocol in one place. Then configure Screaming Frog:

  • Set the user agent to Googlebot Smartphone, since Google indexes the mobile version of your pages.
  • Connect the Search Console and GA4 APIs, and add your XML sitemaps. This lets the crawl flag orphan pages and pages with traffic that are not linked.
  • Run one crawl in text-only mode and a second with JavaScript rendering. You will compare them in step 5.
  • Ask the developer for 30 days or more of server or CDN access logs for step 9.

Step 1: Crawlability

Start with robots.txt. Read it line by line and check the robots.txt report in Search Console (under Settings) to confirm Google fetched the version you think is live. Look for rules blocking CSS, JavaScript or whole product folders.

Then use the crawl to check response codes:

  • 3xx: internal links pointing to redirects, and any redirect chains. Update the links to point at the final URL.
  • 4xx: broken internal links and the pages linking to them.
  • 5xx: any at all. Cross-check the Crawl stats report in Search Console, which shows host status and the share of responses by type over the last 90 days.

Good looks like: robots.txt blocks only URLs with no search value, internal links resolve to 200 pages directly, and the Crawl stats host status shows no recurring availability problems.

Step 2: Indexation

Open the page indexing report in Search Console and read every "not indexed" reason, not only the totals. The ones I look at first:

  • Crawled, currently not indexed: Google saw the page and decided it was not worth indexing. If money pages are here, the page is thin, duplicated, or weakly linked.
  • Discovered, currently not indexed: Google knows the URL but has not crawled it. On large sites this often points to crawl being spent elsewhere.
  • Duplicate, Google chose different canonical than user: your canonical tag is being overruled. Find out why before forcing it.
  • Excluded by noindex tag: confirm every URL here is meant to be noindexed.

Next, check the XML sitemaps. Every URL in them should return 200, be indexable, and carry a self-referencing canonical. Each file can hold up to 50,000 URLs or 50MB uncompressed. Google uses lastmod when it is consistently accurate and ignores priority and changefreq.

In Screaming Frog, filter for canonicalised URLs, noindex URLs and pages where the canonical points to a non-200 URL. Spot-check important templates with URL Inspection to see the Google-selected canonical.

Good looks like: the number of indexed pages is close to the number of pages you want indexed, and none of your revenue pages appear under a "not indexed" reason.

For Bombay Shaving Company, crawl and indexation fixes alongside long-tail landing page clusters were part of the work that produced 3x qualified leads. On Shopify sites the usual culprits are tag pages, filter parameters and duplicate product paths, which I go through in the Shopify indexation bloat post.

Step 3: Site architecture

Check crawl depth in Screaming Frog. In my experience, pages that earn revenue should be within three clicks of the homepage; deeper than that, they tend to get crawled less and linked less.

Run Crawl Analysis to find orphan pages: URLs in sitemaps, Search Console or GA4 that nothing links to. Link them or remove them.

Then look at URL patterns. Parameters, session IDs, sort orders and stacked filters can multiply one page into hundreds of crawlable URLs. Decide which filtered views deserve their own static page and keep the rest out of the crawl. I explain how I make that call for online stores in the category page SEO playbook.

Good looks like: a shallow, logical hierarchy, no orphaned money pages, and a crawl that finishes with roughly the number of URLs you expect.

Sort the crawl by unique inlinks. Footer pages like policies usually top the list. Your priority pages should be close behind.

  • Check anchor text on links to key pages. Descriptive anchors help; "read more" and "click here" do not.
  • Find internal links with rel="nofollow" and remove the attribute unless there is a real reason.
  • Make sure menus and related-content blocks use plain <a href> links, not click handlers on divs or buttons.

Good looks like: your highest-value pages are among the most linked internally, with anchors that describe what is on the page.

Free template

The Technical SEO Audit Template

A quarterly audit your in-house team can run, log and act on without an agency

  • Ten audit areas, each with a checklist and a pass standard
  • The exact Screaming Frog tabs and Search Console reports to open
  • Core Web Vitals thresholds and how to read PageSpeed Insights

Free. Instant access. No spam, unsubscribe any time.

Step 5: JavaScript rendering

Compare the text-only crawl with the JavaScript crawl. Screaming Frog shows raw and rendered HTML side by side, and flags content and links that only appear after rendering. Then use URL Inspection on key templates and view the crawled page HTML and screenshot.

Check that the title, meta robots, canonical, main content and internal links are all present in the rendered HTML. I prefer them in the raw HTML as well, because it removes a dependency on rendering. Watch for a canonical or meta robots tag that changes between raw and rendered versions, and for products or articles that only load on scroll or click.

Good looks like: raw and rendered versions agree on every indexing signal, and the content you want ranked is visible in what Google renders.

Step 6: Core Web Vitals

Core Web Vitals are assessed on real user data at the 75th percentile. The "good" thresholds are:

  • LCP (Largest Contentful Paint): 2.5 seconds or less.
  • INP (Interaction to Next Paint): 200 milliseconds or less.
  • CLS (Cumulative Layout Shift): 0.1 or less.

Start with the Core Web Vitals report in Search Console, which groups similar URLs, so you can see which templates fail. Then open a sample URL in PageSpeed Insights. The top section is field data from the Chrome UX Report; the Lighthouse section below is a lab test. Fix based on field data and use the lab test and Chrome DevTools to diagnose.

The usual fixes: do not lazy-load the hero image, give it fetchpriority="high", set dimensions on images and embeds, reserve space for banners, and cut or defer third-party scripts that block the main thread. Speed work pays off beyond rankings. At your-space, cutting load times by 60% came with a 40% drop in bounce rate.

Good looks like: every main template passes all three metrics on mobile field data.

Step 7: Structured data

Test key templates in the Rich Results Test for Google features and the Schema Markup Validator for general validity. Check the structured data reports in Search Console for errors across the site.

The basics: Organization on the homepage, BreadcrumbList on inner pages, Product on product pages, Article on blog posts. Markup must match what is visible on the page. Do not add markup expecting a rich result Google no longer shows for most sites; FAQ rich results, for example, are limited to well-known government and health websites.

Good looks like: no errors on the types that matter to you, and every marked-up value visible on the page.

Step 8: Hreflang, only if you serve more than one market

Skip this step if you have one language and one market. If you run, say, India and UK versions, check the hreflang tab in Screaming Frog for:

  • Return links: if page A references page B, page B must reference page A.
  • Valid codes: ISO language codes with optional region, such as en-in or en-gb. en-uk is invalid.
  • A self-referencing entry on each page, and an x-default where you have a fallback.
  • Targets that return 200 and are canonical, not redirected or noindexed.

Good looks like: zero hreflang errors in the crawl and the right country version ranking in each market.

Step 9: Log file analysis

Logs show what Googlebot actually requests. Load them into the Screaming Frog Log File Analyser and verify Googlebot hits by reverse DNS, since many bots fake the user agent.

Look at where crawl goes: what share of Googlebot requests land on indexable 200 pages, how many go to parameters, redirects or 404s, and how recently your key pages were last crawled.

Good looks like: most Googlebot requests hit indexable pages, and your priority pages are crawled regularly.

Turn findings into a fix list

An audit only matters if fixes ship. For every issue I record the affected URL count, one example URL, the fix, the owner, and how we will verify it. Then I sort into three groups:

  1. Blocking: stops revenue pages being crawled or indexed. Fix first.
  2. Leaking: wastes crawl or link equity, such as redirect chains, parameter sprawl and orphans.
  3. Polish: markup errors and minor speed issues on low-traffic templates.

To confirm fixes worked, re-crawl and use Screaming Frog's crawl comparison, click "Validate fix" on the relevant Search Console issue, and request indexing for key URLs through URL Inspection. Core Web Vitals field data is a rolling 28-day window, so allow about a month before judging a speed fix.

The short version

  • robots.txt reviewed and confirmed in Search Console
  • No internal links to redirects, 4xx or 5xx
  • Page indexing report reasons reviewed, revenue pages indexed
  • Sitemaps contain only 200, indexable, canonical URLs
  • Money pages within three clicks, no orphans
  • Priority pages among the most internally linked
  • Raw and rendered HTML agree on indexing signals
  • LCP, INP and CLS pass on mobile field data
  • Structured data valid and matching visible content
  • Hreflang reciprocal and valid, if used
  • Logs show Googlebot focused on indexable pages
  • Every finding has an owner and a verification step
Abhishek Dwivedi
Written by Abhishek Dwivedi

Founder of AD Search Studio and a Senior SEO Lead with 7 years in search, running SEO and AEO/GEO programmes across D2C, B2B SaaS and banking brands.

View Abhishek's portfolio

FAQ

Commonquestions.

Still unsure? Ask us directly.

Get in touch

Run a full audit once a year and before any migration, redesign or platform change. In between, a monthly crawl and a look at the Search Console page indexing and Core Web Vitals reports will catch most new problems early.

Free AI search audit

See how Googleand AI see you.

Send your site and our founder reviews it personally. You get a clear, prioritised read on what is holding your search growth back.

  • Technical health: crawl, indexation and speed issues that cap growth
  • AI visibility: how ChatGPT, Perplexity and AI Overviews describe your brand
  • Content gaps: the high-intent queries competitors win and you don't
  • The three fixes worth the most revenue, in order

Free, no obligation. Your details are only used to reply to you.

Talk to the founder

Ready to bethe answer?

Tell us what you want to rank for and where growth has stalled. We will tell you straight whether search is the right lever, and what it is worth.

Abhishek DwivediFounder and Lead Strategist, AD Search Studio

Abhishek Dwivedi