Website Audit Checklist: Fix What Google Can't See

Traffic is flat for months. Search Console shows no penalty, no manual action, nothing. The cause is usually structural: a technical error sits between Google and your pages, one the browser never reveals. A website audit finds those errors in a single pass.

What a website audit actually checks

A website audit is a systematic check of every layer that decides whether Google can find, read, and rank a page: crawl access, index status, on-page signals, speed, and mobile delivery.

Google runs that pipeline in order: crawl, index, rank. Googlebot fetches the URL, the index decides whether the page gets stored, and ranking systems score what survived the first two gates. A page that never gets crawled has no chance at any later step, no matter how good the content is. Google documents this model publicly in its How Search Works guide.

Search Console reports outcomes, rarely causes. It tells you clicks dropped; it does not flag the Disallow: line that started blocking your blog in March. The audit answers why.

A technical SEO audit narrows the scope to the layer visitors never see: crawler access, index status, redirect logic, rendering. The expensive errors live there.

Crawlability and indexability

These errors keep entire sections out of search results. If parts of your site never appear on Google, start here.

Step 1 — Read your robots.txt

1. Open https://yourdomain.com/robots.txt in a browser.
2. Read every Disallow: line. A line reading Disallow: / blocks the entire site. Disallow: /blog blocks a whole section.
3. Confirm the Sitemap: line points to an XML sitemap that returns a 200 status.

Blocked an important path by mistake? Correct the line, then re-request crawling via URL Inspection in Search Console.

Step 2 — Hunt for stray noindex tags

1. Open your homepage and press Ctrl+U (or Cmd+Option+U) to view the HTML source.
2. Search for noindex using the browser's find box.
3. A <meta name="robots" content="noindex"> tag in the <head> removes the page from Google's index outright. These tags commonly survive from staging copies and test deploys.

Delete the tag, then request indexing through the URL Inspection tool. The index-recovery playbook is in our guide to a website not showing on Google.

Broken internal links belong in the same pass. Screaming Frog's free tier (500 URLs) lists every 404; point them at live pages or remove the links. No XML sitemap? Generate one and submit it via Search Console, per the XML sitemap and robots.txt guide.

On-page signals: titles, descriptions, and headings

Three thresholds come straight from public guidance and search-results behavior:

• Title tag: keep it at or under 60 characters; results truncate past that, cutting off your pitch mid-word. Every page needs a unique title that names its main topic.
• Meta description: around 155 characters. Google rewrites them often, but a written summary beats whatever gets auto-extracted from the page body.
• Headings: exactly one <h1> per page, with <h2> and <h3> dividing sections in the order the content presents them.

The duplication check runs in a browser: search site:yourdomain.com in Google and scan the result titles. The same string on twenty pages means templates ship identical titles everywhere, and Google picks which one to show, or none.

Page speed and Core Web Vitals

Google's published thresholds rate LCP above 2.5 seconds as "needs improvement" and above 4.0 seconds as "poor." LCP measures how long the largest visible element takes to render.

Step 3 — Get your LCP number

1. Open PageSpeed Insights.
2. Paste your homepage URL and run the report.
3. Read the LCP value in the Core Web Vitals section and note the flagged element. It is usually the hero image.
4. Check the asset's size in DevTools' Network tab: reload, sort by size.

Fixes that move LCP: compress and resize the hero image to its display dimensions in WebP, add fetchpriority="high" to it, and defer every script the first paint does not need. The full breakdown is in our technical post on slow homepages and rankings.

Mobile, HTTPS, and site-wide hygiene

• Every template ships a viewport meta tag. Pages without one render desktop-width on phones.
• One embedded http:// image triggers the browser's not-secure warning on an https:// page.
• Redirects take one hop: old-url to new-url directly, never through an intermediate stop. Chains waste crawl budget and add latency.
• Each page's canonical tag points at itself unless it deliberately duplicates another URL.
• Meaningful images carry descriptive alt text; decorative ones carry alt="".

Fix your findings, then re-run the audit

Work in order: crawl and index errors first, since they make the site invisible; on-page fixes second; speed third. Re-run the audit after each batch so the change is measurable.

That is how to audit a website without paying an agency: paste your domain into a free website audit tool, read the prioritized error list, and start with whatever blocks crawling or indexing. SiteGrade's free homepage audit runs 50+ checks across the error classes in this list (crawl access, titles and headings, speed, HTTPS, redirects) and ranks them by impact.

A free website audit tool gives you the same first-pass data an agency bills for. The sequence above sets the order.

Run it once today, fix in that order, and re-run to confirm.