To do an SEO audit, you work through a fixed process: define the scope, gather your tools, crawl the site, pull Search Console, run the technical, content, keyword, internal-linking, off-page, and AI-visibility passes, then analyze everything and prioritize the fixes. The first seven steps collect data. The last step, turning that data into a ranked plan, is the actual audit.

That distinction is where most DIY audits go wrong, and where paid SEO services earn their fee. People run a tool, get a score, and stop, mistaking a pile of findings for a diagnosis. A real audit reads the findings together and answers one question: of everything wrong here, what is costing the most visibility and pipeline, and what do we fix first? Below is the exact process I run on B2B SaaS sites, in order.

If you want the full inventory of individual items to check inside these steps, keep the complete SEO audit checklist open alongside this. That is the what; this is the how.

Step 1: Define the scope and pull a baseline

Before you touch a tool, decide what you are auditing and why. A full-site audit, a technical-only pass, and a “why did this section stop ranking” investigation are different jobs with different scopes. Write down the goal and the questions you need answered.

Then pull a baseline: current organic traffic, current rankings for your priority terms, and the number of indexed pages. You need a starting line, both to spot what is already broken and to measure whether your fixes worked later. Skipping the baseline is how audits become unfalsifiable.

Step 2: Gather your tools and access

An audit runs on data from a few sources, so line them up first:

  • A crawler (Screaming Frog is the standard) to map the site the way a bot sees it.
  • Google Search Console for indexation, queries, and how Google actually treats your pages.
  • A keyword and rank tool (Ahrefs or Semrush) for keyword rankings, page-level traffic estimates, and competitor analysis. This is the one kind of data Search Console cannot give you, because it only covers pages you own. Seeing which terms competitors rank for, and which of their pages actually earn the traffic, is most of real competitor analysis, and it is the best and one of the only ways to do it properly.
  • Analytics for traffic and, ideally, conversions by channel.
  • PageSpeed Insights or the Core Web Vitals report for performance.

Confirm access to all of them before you begin. Half the stalls in a DIY audit come from discovering you cannot get into Search Console on day two.

You can run a genuine audit on free tools, which is the honest answer to how to do an SEO audit for free, and how to do an SEO audit for beginners is the same process on the same free stack. The best free SEO audit tools get you a long way: Search Console and PageSpeed Insights are free, and Screaming Frog crawls up to 500 URLs at no cost, enough for most small sites. Even keyword research has a free floor: Search Console’s own query report, Google Keyword Planner, and the free tiers of the paid tools tell you a lot about what you rank for and what is worth targeting before you spend anything. Paid tiers and tools like Ahrefs or Semrush add depth and scale, but cost is not what stops a beginner from starting. Judgment is.

Step 3: Crawl the site

Run a full crawl to see every URL the way a search engine does. A crawl maps status codes, redirects, titles, meta descriptions, headings, and the internal link graph, and it immediately surfaces the mechanical problems: broken links, redirect chains, orphan pages, missing titles, and duplicate content. Export it. This is the raw material for the technical and on-page passes.

Step 4: Check indexation and technical health

Now cross-reference the crawl against Search Console, because the crawl tells you what exists and Search Console tells you what Google did with it. This is where the highest-leverage technical problems hide:

  • What is actually indexed versus what you want indexed.
  • Pages sitting in crawled, currently not indexed, which is a quality verdict, not a bug.
  • Core Web Vitals and mobile usability.
  • JavaScript rendering that hides content and links from crawlers and AI bots.
  • Schema markup validity on the templates that warrant it.

On one cybersecurity SaaS account, this exact cross-reference surfaced roughly a thousand pages stuck in crawled-but-not-indexed limbo. A single robots.txt fix cleared a few hundred immediately. You do not find problems like that from a score out of 100; you find them by reading the crawl and Search Console together.

Step 5: Audit content, keywords, and intent

With the foundation mapped, judge whether the content deserves to rank. Start with who it is for. You cannot audit keywords or intent in a vacuum: “does this match intent” only means something relative to the persona you are selling to, so pin down the ICP and the questions they actually ask before you grade a single page. Then inventory every indexable page and give each a verdict: keep, refresh, consolidate, or prune, and layer on the keyword view:

  • Intent match: does each page match the search intent and page type Google rewards for its target term, for the persona you are actually selling to?
  • Cannibalization: are two pages fighting for the same keyword and capping each other? Consolidate them.
  • Coverage: where are the bottom-of-funnel, buyer-intent gaps, and the striking-distance terms you already rank for at positions 5-20? This is where a rank tool like Ahrefs or Semrush earns its keep, surfacing the terms competitors rank for and you do not.
  • Manual SERP comparison: for your money terms, no tool makes the call for you, so put your page beside the page actually ranking and the competitor page next to it, and compare them by hand. Do it for your core pages and your blog posts alike: what does the ranking page cover, and in what depth and format, that yours does not?
  • Information Gain: is the content something only you could publish, or commodity synthesis that core updates now filter out?

On a creator-tools account, this pass scored roughly 900 pages and flagged the majority for consolidation or deletion. A content audit’s job is often to find the dead weight, not to recommend more posts.

Step 6: Check internal linking and off-page authority

Two quicker passes that punch above their weight. For internal linking, check for orphan pages, hub-and-spoke coverage between pillars and spokes, and money pages buried too many clicks deep. For off-page, read the backlink profile’s quality and trend, and check where the brand is mentioned across the third-party sources, G2, Reddit, industry press, that shape both search and AI answers.

Step 7: Test AI-search visibility by hand

This is the step guides written before 2024 do not have, and it is now half the audit. Ranking on Google is no longer the same as being found, because buyers increasingly get their answer from ChatGPT and Google’s AI Overviews without a click. There is no single dashboard for this yet, so you test it manually:

  • For each money keyword, check whether you are named, only cited, or absent in ChatGPT, Google AI Overviews, Perplexity, and Gemini.
  • Note where competitors get named and you do not, the AI-era version of a keyword gap.
  • Flag which ranking terms now trigger an AI Overview that eats the click, so you know where traffic is quietly draining.

Do this by hand for your top terms and you learn more about your real visibility than any legacy rank tracker will tell you.

Step 8: Analyze, prioritize, and write the plan

Everything so far was data collection. This step is the audit. Read the findings from all seven passes together, because they interact, a content problem is often really an indexation problem, and rank them by impact on pipeline versus effort to fix.

Knowing how to do SEO audits well comes down to this step, and it is also how to prepare SEO audit report output people actually act on. The result is not a 200-row export. It is a short, prioritized plan: the handful of high-leverage fixes first, quick wins separated from big projects, each tied to why it matters. Quick wins are not a fixed category; they are whatever is high-impact and low-effort on this particular site, sometimes a set of URLs or redirects, sometimes a few core pages re-optimized, sometimes on-page tweaks or a single technical fix. It is different every time. That prioritization is the entire value of the exercise, and it is the one part a tool cannot do for you.

Running the tools is data collection. The audit is the judgment call about what to fix first. Confuse the two and you get a report nobody acts on.

The reason to sequence by pipeline is that it changes what you work on. On a legal SaaS account I run, the audit pointed the plan at bottom-of-funnel comparison and category pages, and that focus drove organic to the single largest demo channel in the business: 362 influenced demos in six months. A checklist would have flagged a hundred things. The audit picked the few that moved revenue.

How to do an SEO audit FAQ

Can I do my own SEO audit?

Yes. With a crawler, Search Console, analytics, and this process, you can run a competent audit and catch the obvious problems yourself. The ceiling is judgment: reading the data from every pass together and knowing which few findings actually move pipeline for your site, then sequencing them. The tools gather symptoms; prioritization is the expert part. Run it yourself to find the low-hanging fruit, and bring in help when the list is long and you are not sure what to fix first.

Can ChatGPT do an SEO audit?

No, not a real one. ChatGPT cannot crawl your site, read your Search Console, or rank fixes by their impact on your pipeline, which are the core of an audit. It is genuinely useful for pieces of the work, drafting a title tag, explaining a crawl error, sanity-checking a hypothesis, but it cannot run the process or make the prioritization call. Treat it as an assistant inside the steps, not a replacement for them.

How to do an SEO audit for a job interview?

Learning how to do an SEO audit for a job assignment is the same process, scoped down to what you can see without their logins. You will not have their Search Console, so work from what is public: crawl the site with Screaming Frog’s free tier, sanity-check indexation with site: searches, read the intent and titles on their money pages, test whether they are named in ChatGPT and AI Overviews for their category, and spot-check Core Web Vitals in PageSpeed Insights. Then do the part that actually gets you the offer: don’t present a list of everything wrong. Pick the two or three findings that would move the most pipeline, say what you would fix first and why, and what access you would need next. Interviewers are testing prioritization and judgment, not whether you can run a crawler.

How much does an SEO audit cost?

A self-run audit costs your time. A professional, engagement-grade audit typically runs from about $1,500 for a freelancer to $25,000 and up for a large agency, depending on site size and depth. I break the tiers down in what an SEO audit costs, and the wider channel pricing in how much SEO costs, and my own fixed-fee audit sits deliberately in the middle: agency-grade depth without the retainer.

What is the 80/20 rule for SEO?

The 80/20 rule is the observation that roughly 20% of your pages and fixes drive about 80% of your results. It is why the prioritization step matters more than the checklist: a good audit finds the vital 20% worth funding and the long tail worth pruning, so you spend effort on the handful of things that move rankings and pipeline instead of grinding evenly through every finding.

How long does an SEO audit take?

A focused self-run pass on a small site can take a few hours. A thorough audit of a real B2B SaaS site usually takes a couple of weeks, because the value is in analysis, cross-referencing the crawl against Search Console, testing AI visibility by hand, and building the prioritized plan, not just collecting the data. Site size and how much is already broken move the number.