How to do an SEO audit: a step-by-step guide for 2026

What an SEO audit actually checks

An SEO audit reviews three layers of your site: technical health, on-page signals, and off-page factors. Each layer affects rankings differently and requires different tools and fixes.

The technical layer covers how Google crawls and indexes your site. If your pages are blocked by robots.txt, returning server errors, or caught in redirect loops, Google cannot index them regardless of how good the content is. Technical issues are the most urgent to fix because they affect everything above them.

The on-page layer covers the content signals Google uses to match your pages to search queries. This includes title tags, meta descriptions, heading structure, keyword usage, content depth, internal links, and structured data. On-page problems are the most common: most sites have dozens of pages with missing or duplicate title tags, weak meta descriptions, and thin content that underperforms relative to competitors.

The off-page layer covers backlinks, brand signals, and E-E-A-T factors. A backlink audit identifies toxic links pointing at your site, lost links from previously referring domains, and gaps relative to competitors ranking above you. This layer matters most for competitive keywords where technical and on-page factors are similar across the top-ranking pages.

For tools to support each stage, the best SEO audit tools guide compares options across each category by site size, budget, and use case.

Step 1: Crawl your site

A crawl is the foundation of every technical audit. It replicates what Googlebot does when it visits your site and produces a complete picture of every URL, its status code, its on-page data, and any technical issues attached to it.

To crawl your site, use Semrush Site Audit, Ahrefs Site Audit, or Screaming Frog. Set the crawler to follow all links including subdomains if relevant, and configure it to render JavaScript if your site uses a JavaScript framework. Crawl depth matters: a shallow crawl misses pages buried more than three clicks from the homepage, which are often the pages with the most issues.

Once the crawl completes, review these outputs first:

• 4xx errors. Pages returning a 404 or 410 status code. Any that have inbound links need a redirect or need the linking page updated.
• 3xx redirect chains. Redirects passing through more than one hop waste crawl budget and dilute link equity. Flatten all chains to direct 301s.
• 5xx server errors. Server errors during a crawl usually indicate capacity issues or misconfigured hosting. Fix these before anything else.
• Blocked URLs. Pages in your sitemap that are blocked by robots.txt. Either update the sitemap to remove them or update robots.txt to allow crawling if the block was unintentional.
• Duplicate content. Pages with the same or near-identical content. These need canonical tags pointing to the preferred version.

Export the crawl data to a spreadsheet and sort by issue type. Most audits surface more issues than can be fixed in one sprint. The crawl data informs the prioritisation stage at the end.

Step 2: Check indexation and coverage

A crawl tells you what URLs exist on your site. Search Console tells you which ones Google has actually indexed. The gap between the two is where many ranking problems live.

Open the Coverage report in Google Search Console. Review four categories:

• Valid pages. These are indexed and eligible to rank. Check the count matches your expectations. If it is significantly higher than your actual page count, you likely have duplicate or auto-generated URLs in Google's index.
• Valid with warnings. Indexed but flagged, usually for duplicate content or alternate pages. These need review.
• Excluded. Pages Google chose not to index or that you explicitly excluded. Review the reasons. Noindexed pages that should rank need the noindex directive removed. Pages excluded for duplicate content need canonical tags.
• Errors. Pages Google tried to crawl and could not. These are crawl errors and need immediate attention.

Cross-reference your sitemap against the Coverage report. Every priority page should be in the sitemap and indexed. Pages that have been live for more than six weeks and are still not indexed either have a crawl error, are blocked, or have a content quality issue that Google is choosing not to index.

Step 3: Audit on-page fundamentals

Once you confirm your priority pages are indexed, audit the on-page signals that tell Google what each page is about and how relevant it is to specific queries.

Work through these checks for every priority page:

• Title tag. Is it present, unique, and between 50 and 60 characters? Does it contain the primary keyword close to the start? Duplicate title tags across multiple pages are a common source of keyword cannibalisation.
• Meta description. Is it present and between 145 and 160 characters? Does it contain the primary keyword and a reason to click? Missing meta descriptions mean Google generates one from the page content, which is usually less effective.
• H1 tag. Is there exactly one H1 per page? Does it match or closely relate to the title tag? Pages with no H1 or multiple H1s send mixed signals about their primary topic.
• Content depth. Does the content cover the topic more thoroughly than the top-ranking competitors for the target keyword? Use Surfer SEO to compare your content score against the top 20 results for each keyword.
• Internal links. Does the page receive internal links from other relevant pages on your site? Is the anchor text descriptive rather than generic? Pages with no internal links are harder for Google to discover and harder for it to determine their authority.
• Structured data. For pages where structured data applies (articles, products, FAQs, how-tos), is the schema markup present and valid? Use Google's Rich Results Test to check each page.

Rank Math handles this check automatically for WordPress sites, flagging issues within the editor. For non-WordPress sites, export title tags and meta descriptions from your crawl and review them in a spreadsheet sorted by page type and priority.

Step 4: Review backlink health

A backlink audit reviews the quality, relevance, and volume of links pointing to your site from external domains. It also identifies lost links that previously passed authority to your site and now return a 404.

Open Ahrefs or Semrush and pull a full backlink report for your domain. Work through these checks:

• Referring domain count and trend. Is the number of unique referring domains growing, flat, or declining? A declining trend often precedes a ranking drop by several months.
• Lost backlinks. Pages that used to link to your site but now return an error. If the linking page still exists, contact the publisher to request the link is restored or updated.
• Toxic links. Links from spammy, irrelevant, or penalised domains. Ahrefs and Semrush both flag these with a spam or toxicity score. Build a disavow file for any links you cannot get removed manually and submit it to Google Search Console.
• Anchor text distribution. A healthy backlink profile has a mix of branded, generic, and keyword-rich anchor text. Over-optimised anchor text (too many exact-match keyword anchors) is a manual penalty risk.
• Competitor backlink gaps. Use the Link Intersect or Backlink Gap tool in your chosen platform to find domains linking to your competitors but not to you. These are the highest-priority link building opportunities.

Step 5: Assess content quality

A content audit identifies which pages are performing, which are declining, and which should be updated or removed. Pull organic traffic data for every page from Google Analytics and sort by sessions over the past 12 months.

Flag pages in three categories:

• Declining traffic. Pages that ranked well 12 months ago but have lost traffic consistently. These need a content update: expand the topic coverage, update outdated information, and refresh internal links. Check whether a Google core update affected these pages specifically.
• Zero traffic. Pages that have never attracted organic traffic. Either they target keywords with no search volume, they have a technical issue preventing indexation, or the content quality is too thin to rank. Decide whether to update, consolidate, or remove each one.
• Cannibalisation. Multiple pages targeting the same keyword. Google struggles to determine which one to rank, which splits authority and reduces the ranking potential of both. Consolidate into a single stronger page using a canonical tag or a 301 redirect from the weaker URL.

For content that needs updating, Surfer SEO's Content Editor gives you a keyword-by-keyword breakdown of what the top-ranking pages include that yours does not. It is the fastest way to identify specific gaps rather than guessing what to add.

Step 6: Prioritise fixes by impact

An audit that produces a list of 200 issues without prioritisation is not useful. You need to know which fixes will move rankings most, given the time and resource available to implement them.

Score every issue on two dimensions: traffic impact and fix effort. A broken canonical tag on your highest-traffic page has high impact and low effort. A missing image alt text on a low-traffic page has low impact regardless of effort. Fix in this order:

1. Critical technical issues. Crawl errors, indexation blocks, server errors, and broken sitemaps. These prevent Google from accessing your content and need fixing before anything else.
2. Duplicate content and canonical errors. Duplicate title tags, missing canonicals, and keyword cannibalisation. These dilute authority across multiple pages and suppress rankings across the board.
3. On-page gaps on priority pages. Missing or weak title tags and meta descriptions on your highest-traffic and highest-potential pages. Quick wins with measurable click-through rate improvements.
4. Content updates on declining pages. Pages that used to rank well and have declined. A targeted content update often recovers lost rankings faster than building a new page from scratch.
5. Backlink toxicity. Disavow clearly toxic links. This is a lower priority than technical and on-page fixes for most sites unless there is evidence of a manual penalty.
6. Content consolidation. Merge thin, low-traffic pages into stronger ones. This takes longer to show results but improves the site's overall topical authority over time.

Document every issue in a spreadsheet or a project management tool like Notion or Trello. Assign each fix to an owner with a deadline. An audit without an owner for each action item produces no results. Using a structured SEO audit report template makes this easier, particularly when communicating findings to a client or developer who was not involved in the audit itself.

What this means for your site health

Running an SEO audit is the clearest way to understand why your site ranks where it does and what would need to change to rank higher. The six-step process in this guide applies to sites of any size. The tools differ depending on your budget and site complexity, but the sequence is the same whether you are auditing a 10-page brochure site or a 100,000-page e-commerce store.

The most important habit is acting on what you find. Audits that sit in a folder while fixes remain unscheduled produce no improvement. Set a follow-up date for each fix, track whether rankings improve after implementation, and schedule the next audit before you close the current one.

For an ongoing view of your site's health between audits, the best SEO audit tools guide covers the monitoring and alerting features in Semrush and Ahrefs that catch new issues before they become ranking problems. Auditing is most effective when it connects to a standing monitoring workflow rather than a quarterly scramble to understand what went wrong.