SEO Audit Template: What to Check and How
A working SEO audit template covering access, indexing, content, links and speed, with what to check in each layer and how to tell pass from fail.
A useful SEO audit checks five layers in order — access, indexing, content, links, speed — because each one caps what the next can achieve. This template lists what to check in each layer, what a pass looks like, and where to look for the answer.
It is the working checklist version of the site's technical SEO guide, built to be run against your own site rather than read as theory.
| Layer | What it answers | Tool |
|---|---|---|
| Access | Can crawlers reach the site at all? | robots.txt, server logs, DevTools |
| Indexing | Is what they can reach actually stored? | Search Console |
| Content | Does each page answer a real query well? | Manual review, search results |
| Links | Does authority reach the pages that need it? | Site crawl, backlink tool |
| Speed | Does the page load and respond fast enough? | Search Console, PageSpeed Insights |
في هذه الصفحة
How do I check the access layer?
Confirm nothing is stopping a crawler before you look at anything else. Fetch /robots.txt and read every rule; a wildcard Disallow: / left over from staging, or a rule that accidentally blocks an entire section, is the single most damaging finding an audit can turn up, because it silently caps everything below it [3].
Check the same file for AI crawler user agents specifically. A firewall or hosting default that blocks scraping bots often blocks the retrieval crawlers behind AI answers too, with no error and no obvious symptom beyond the brand never showing up in that channel. Our guide to robots.txt for AI crawlers covers exactly which agents to name and allow.
Confirm server responses are healthy: no unexpected redirect chains, no 5xx errors on pages that should return 200, and HTTPS enforced everywhere with no mixed-content warnings. A slow or flaky server under crawl load fails this layer just as surely as a robots.txt block does.
How do I check the indexing layer?
Pull the Page indexing report in Search Console and read the full breakdown, not just the total indexed count. Sort the excluded URLs by reason: blocked, noindexed, duplicate, or simply not yet crawled. Each reason points at a different fix, which our guide to indexing problems breaks down in more depth than fits here.
Check canonical tags next. Every page should self-canonicalize unless there is a specific reason it should not, and a canonical pointing at the wrong page, a parameter version, or a page that no longer exists is a common, quiet cause of the wrong URL ranking.
How do I check the content layer?
Read the top twenty pages by traffic potential the way a search engine or an assistant would: does the page answer the question in the title within the first few sentences, or does it take three paragraphs to get there? Pages that bury the answer lose to pages that lead with it, in both classic rankings and AI citations.
Check for thin or duplicate content next, especially on any page generated from a template: product variants, location pages, or filtered category pages. A page that reads as a near-copy of another with a few words swapped is a page competing against your own site rather than for it.
List your top twenty pages by current or potential traffic
Start where the impact is largest, not where the problems are easiest to find.
Read each one as a first-time visitor would
Does it answer the query in the title within the first few sentences?
Check for a table, a stat or a quote on each page
Pages with none of these are the ones assistants pass over for a competitor's page that has them.
Flag near-duplicates for merging or differentiation
Two pages saying the same thing compete with each other more than with a competitor.
How do I check the links layer?
Two directions matter: internal links that spread authority to the pages that need it, and external links that bring authority into the site at all. Check internal links first, because it is the cheaper fix and entirely within your control.
Crawl the site and count inbound internal links per page. Any page more than three clicks from the homepage with few or no internal links pointing at it is effectively orphaned, competing poorly for crawl attention and ranking signal alike regardless of how good its content is.
For external links, a backlink tool will show what points at the site today; the audit question is whether any of it looks manipulated or comes from sites with no real relevance, which is worth a cleanup, and whether the strongest pages by traffic actually have any external links at all, which is often the real gap on a newer domain.
How do I check the speed layer?
Open the Core Web Vitals report in Search Console and read the field data, not a single lab score. A page passes when Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift are all good for at least 75% of real visits [2]. Our Core Web Vitals guide has the full fix order once you know which metric is failing.
Check this by template, since one slow layout usually explains most of a site's failing URLs at once. Identify the template with the most traffic that fails, fix it first, and confirm the fix in the field data about four weeks later rather than trusting a lab score the day you ship it.
How do I turn the findings into a plan instead of a list?
Score each finding on two axes: how much traffic or revenue it touches, and how hard it is to fix. A robots.txt block on a high-traffic section is both high-impact and low-effort, so it goes first. A content rewrite across two hundred thin pages is high-impact but high-effort, so it becomes a project with a timeline rather than a single ticket.
Write every finding down with an owner and a page or template it applies to, not as a general statement about the site. "Fix slow LCP" is not actionable; "hero image on the product template is 2.4MB and loads without a size attribute" is.
Group findings by template rather than by individual URL wherever the cause repeats. A single fix to a shared header, a shared product template or a shared redirect rule often clears dozens of findings at once, and listing them as one ticket instead of dozens keeps the plan short enough that someone actually works through it. A fifty-line spreadsheet gets read; a five-hundred-line one gets filed away.
Set a date to re-check the same five layers after the highest-priority fixes ship, rather than assuming a fix worked because it was deployed. Indexing and rankings both lag a change by days to weeks, and the only way to know a fix held is to run the relevant slice of this template again against the same pages once enough time has passed.
If running all five layers yourself is more than your team has time for, that full pass, ranked and written up the same way, is what our SEO audit service delivers, with the bias disclosed plainly: it is our own service, and every step in this template works just as well run in-house.
أسئلة يطرحها الناس
How long does a full SEO audit take to run?
A DIY pass through this template takes a focused day or two for a small site, longer for one with thousands of URLs or several languages. An agency audit usually takes one to two weeks because it adds log analysis and manual review on top of the automated checks.
Do I need paid tools to run this audit?
No. Search Console, a free crawler with a URL limit, and a browser's own DevTools cover most of this template. Paid tools save time on large sites by removing crawl limits and automating the cross-referencing, but they do not check anything a free stack cannot check by hand.
Which section of the audit matters most?
Access, because every other layer depends on it. A perfectly optimized page that a crawler cannot reach or is told not to index will not rank regardless of what the content or link layers find.
How often should we re-run a full audit?
Once a year for a stable site, or after any major change: a migration, a redesign, a platform switch, or a period of unexplained traffic loss. Smaller spot checks on the layers most likely to drift, like indexing and speed, are worth doing monthly.
Should the audit include AI search visibility, or is that separate?
Include it. Whether AI crawlers can reach and quote your pages depends on the same access and content quality checks as traditional search, so it belongs in the same pass rather than a separate project.
المصادر
- SEO Starter Guide — Google Search Central (checked 22 September 2026)
- Understanding Core Web Vitals and Google search results — Google Search Central (checked 22 September 2026)
- Robots.txt introduction — Google Search Central (checked 22 September 2026)
Want the five-layer audit run against your real data?
احجز استشارة مجانية مدتها 30 دقيقة مع إبراهيم وجيريمي. ننظر في موقعك أثناء الحديث، وتخرج بخطة مرتّبة حسب الأولوية.
من كتب هذا
Jeremy ZavaCo-founder
Leads growth and client partnerships.
يعمل على: CRO, SEO
راجعه قبل النشر Ibrahem Helal، Co-founder.
المزيد عن هذا
- Technical SEO: the complete checklist for 2026Technical SEO makes sure search engines and AI crawlers can reach, render and index your pages. This checklist covers every layer, in the order to fix it.
- Site Migration SEO: The Full ChecklistA site migration loses traffic when redirects, canonicals or content go missing. This checklist covers the mapping, the launch and the four weeks after.
- Indexing Problems: Why Pages Stay OutPages stay unindexed for a handful of reasons: low quality, a blocked crawl path, a wrong canonical, or Google simply choosing not to bother yet.
- Robots.txt Rules for AI CrawlersA working robots.txt for AI crawlers names GPTBot, ClaudeBot and PerplexityBot explicitly and never blocks the ones your citations depend on.