Technical SEO checklist: every check, in the order that matters

Work a technical audit in one order: can the page be crawled, is it indexed, does it render, is it the only version of itself, is it fast, and is it marked up correctly. Each stage depends on the one before, so a speed project on pages Google never crawls returns nothing. You need three instruments and no more to start: Google Search Console, a site crawler and PageSpeed Insights. Everything below is checkable with those.

Work top to bottom, fix as you go, and record the date you checked each item so the next audit is a comparison rather than a fresh start.

1. Crawling

Check	How to verify
robots.txt serves a 200 and allows what it should	Fetch /robots.txt directly, then use the robots.txt report in Search Console
No blanket Disallow left over from a staging site	Read the file line by line; a single stray slash can block everything
No noindex page is blocked in robots.txt	Google states that a page must not be blocked by robots.txt for noindex to take effect, since the crawler never reads the rule. See block indexing with noindex
Crawl budget is not going on filter and search URLs	Check server logs or the crawl stats report for the URL patterns being requested most
XML sitemaps list only indexable, canonical URLs	Crawl the sitemap and check each URL returns 200 and self canonicalises
Each sitemap is within 50,000 URLs and 50MB uncompressed	Google's stated limits in the sitemap guide; split with a sitemap index above either
Status codes are correct	Crawl the whole site. Google follows up to ten redirect hops and treats 301 as a stronger signal than 302, per its status code documentation
No redirect chains or loops	Point every redirect at the final destination in one hop
One hostname and one protocol	http, https, www and non www should all end at a single version by redirect

2. Indexing

• Read the Pages report in Search Console in full, not just the total. Each exclusion reason is a different problem: discovered but not indexed points at crawl budget or perceived value, crawled but not indexed at quality or duplication, and alternate with canonical at your own canonical choices.
• Check your top twenty revenue pages individually with the URL Inspection tool, which shows the indexed version, the canonical Google chose and the structured data it found.
• Confirm the canonical Google selected matches yours. Where they differ, the page is telling a mixed story.
• Check canonical consistency across signals. Google's canonical documentation calls a redirect and a rel="canonical" annotation strong signals and sitemap inclusion a weak one, asks for absolute URLs, and asks that the signals not contradict each other.
• Look for pages competing with each other on the same phrase. Merge them into the stronger URL and redirect the rest.
• Verify hreflang if you serve several markets: every annotation returns, each page references itself, and the language and region codes are valid.

3. Rendering

Google's JavaScript SEO documentation describes three phases, crawling, rendering and indexing, and explains that pages returning a 200 enter a rendering queue where a headless Chromium executes the JavaScript. The practical checks follow from that.

• View the rendered HTML in URL Inspection and confirm your main content and internal links are present in it.
• Put primary content and links in the initial HTML where you can, since server side rendering helps every crawler.
• Use real anchor elements with href attributes for navigation. Google follows links in href, not click handlers.
• Keep the resources needed for rendering, such as JavaScript and CSS files, crawlable.
• Avoid content that appears only after a user action, since nothing clicks on your behalf.

4. Architecture and internal links

• Every page you want indexed is reachable by links within three clicks of the home page.
• No orphan pages: crawl the site and compare against the sitemap and your content management system list.
• Breadcrumbs are present and marked up, following Google's BreadcrumbList documentation, which requires at least two ListItem entries each with position and name.
• URLs are lowercase, stable, readable and free of session parameters.
• Pagination uses crawlable links. Google's pagination guidance notes crawlers follow links in the href attribute, which makes ordinary pagination the most crawler friendly pattern.
• Faceted and filter URLs are kept out of the crawl unless they have real demand, per Google's faceted navigation guidance.

5. On page basics

• One unique, descriptive title per page, with the main phrase near the front.
• One h1 per page that matches what the page is about, and a logical heading order below it.
• A first paragraph that answers the search outright, because that is what gets quoted.
• Meta descriptions written for the pages that matter, and left to the engine elsewhere.
• Images with real file names, descriptive alt text, modern formats and explicit width and height.
• Internal links from your strongest pages to the pages you want to rank, with descriptive anchor text.

6. Speed and Core Web Vitals

Use field data from the Core Web Vitals report in Search Console for diagnosis, and PageSpeed Insights for a single page. The published thresholds, each measured at the 75th percentile of page loads and segmented by mobile and desktop, are:

Metric	Good	Poor
Largest Contentful Paint	2.5 seconds or less	Above 4.0 seconds
Interaction to Next Paint	200 milliseconds or less	Above 500 milliseconds
Cumulative Layout Shift	0.1 or less	Above 0.25

The fixes that move these most often: compress and correctly size images, serve the hero image without waiting on JavaScript, remove third party scripts you are not using, reserve space for anything that loads late such as ads and embeds, defer non critical JavaScript, and cache properly at the edge.

7. Structured data

• Mark up what the page actually is, and only what is visible on it.
• Validate every template with the Rich Results Test and the structured data report in Search Console before rolling it out across thousands of URLs.
• Product markup belongs on single product pages. Google's merchant listing documentation requires name, image and an offers object with price and priceCurrency, and states that only pages where a shopper can buy are eligible for merchant listing experiences.
• LocalBusiness markup requires at least name and address, per Google's local business documentation.
• Declare your site name with WebSite structured data on the home page, which Google uses when choosing the site name shown in results.
• Keep the markup in sync when the page changes. Stale prices and availability are the most common structured data fault.

8. Security, international and monitoring

• HTTPS everywhere, with a valid certificate and no mixed content warnings.
• A monitoring alert on robots.txt changes, on sitemap errors, and on a sudden drop in indexed pages.
• A pre release checklist for every deployment covering robots.txt, noindex tags, canonical tags and redirects, because most sudden traffic losses are a deployment, not an algorithm.
• A quarterly full crawl, compared against the last one.

This page covers the technical layer. For the tools that run each check and what each kind looks at, see what each kind of SEO audit tool checks. For diagnosing a sudden drop, work through a traffic drop checklist in order.

What gets asked about Technical SEO checklist

How often should I run a technical audit?

Run a full crawl quarterly, read the Search Console reports monthly, and run the deployment checklist on every release. The release check catches more problems than the quarterly audit does.

Which items matter most on a small site?

Indexing, one version of each URL, titles and first paragraphs, mobile speed, and internal links. A small site rarely has a crawl budget problem, so the effort belongs in content coverage and clarity.

Do I need a paid crawler?

Search Console plus a crawler is the minimum. The free tiers of the common crawlers cover a few hundred URLs, which is enough for a small site; above that a paid licence pays for itself in a single audit.

Is robots.txt the way to remove a page from Google?

Use noindex on a URL that stays crawlable. Google is explicit that robots.txt is not a mechanism for keeping a page out of search, and a page blocked there never has its noindex rule read. See the robots.txt introduction.

How long after a fix should I expect to see movement?

Indexing changes appear within days for important pages, especially after requesting a recrawl. Ranking and traffic effects build over weeks, so set the review date a month out and compare against the baseline you recorded.

One more line for the quarterly audit

The checklist ends with monitoring: crawl errors, indexing, speed, structured data. Add one line that sits after it. Ask the three main assistants the questions your key pages answer, and write down whether you are cited, and which pages they trust instead.

A page that passes every technical check can still be missing from those answers, and the audit would not have caught it. AI Knows Us produces this line each quarter for a fixed list of questions, which makes audits comparable.