How to do an SEO audit step by step – technology, business and AI visibility (and what an automated audit misses)

An SEO audit is a structured review of why a website is visible – or not – when the right people search. It answers three questions: does the technical side work, is the content targeting keywords that actually drive business, and is the company visible when buyers ask an AI instead of googling? An automated performance score of 0–100 only answers the first question. This guide walks through how to do an SEO audit step by step and which tools you need – and you can also test how your own website scores on the automated metrics in Google's PageSpeed Insights tool.

What is an SEO audit – and what is it not?

An SEO audit maps how well a website is positioned to rank and convert. It looks at the technical foundation, the relevance of the content to search intent and how the website holds up against competitors in the search results. The goal is a prioritised action list – not a number.

It is worth distinguishing the audit from two things it is often confused with. Keyword research maps which search terms your audience uses and how big the demand is. It is one part of an SEO audit, not the whole thing. Nor is an SEO audit the same as troubleshooting: if the website lost its rankings overnight, you are usually looking for a specific error, not a strategic review. We have collected the most common causes in Website not ranking? The 10 most common causes – start there if the problem is urgent, and do the audit when you want to build for the long term.

A complete SEO audit covers three layers: the technical layer, the business context (which keywords drive revenue, not just traffic) and AI visibility. Most free tools stop at the first layer.

An SEO audit shows where you stand – an SEO strategy is the plan that points the way forward. Do the audit first; the strategy builds on what it shows.

Free technical SEO audit – test your website here

Enter a URL and Google Lighthouse runs a free technical SEO audit of the page: performance, accessibility, Core Web Vitals and nine SEO checks. The results appear after 20–30 seconds.

Test your page's performance

Enter an address and get Google's Lighthouse scores plus a review of the technical SEO checks. Measured for mobile by default.

The checks the tool reports on:

• The page has a title tagdocument-title
• The page has a meta descriptionmeta-description
• Correct HTTP status codehttp-status-code
• The page can be indexed by Googleis-crawlable
• Links have descriptive textlink-text
• Links are crawlablecrawlable-anchors
• Images have alt textimage-alt
• Correct hreflang (language/region)hreflang
• Mobile-friendly viewportviewport-insight

More free tools

What a free SEO audit measures – and what it does not

Free SEO tools take a URL and return a score between 0 and 100 plus a list of technical warnings. They are useful for a quick overview – but it is important to understand what the score is: a weighting of technical signals the tool can read automatically.

What a free score measures well: broken links, missing meta descriptions, images without alt text, pages that load slowly, missing schema markup. These are real issues and worth fixing.

What a free score cannot measure is everything that requires judgement and business context:

• Whether the keywords drive business – a tool sees that you rank, but not whether the term attracts buyers or just the curious.
• Competitor gaps – why a competitor ranks above you for the profitable terms.
• Search intent – whether the content matches what the person actually wants when they search.
• AI visibility – whether the company is mentioned correctly in ChatGPT, Claude or Google's AI Overviews.

In other words: a high score means the technical side is in order. It says nothing about why the business is not growing. That is the difference between troubleshooting a website and analysing a business.

How to do an SEO audit step by step

Here is a sequence that covers all three layers. Work through the steps in order – each one builds on the last.

1. Set the goal and define the business value. Decide what the website should achieve: leads, sales, bookings. Rankings and traffic are means to an end, not goals.
2. Do keyword research. Map the terms your audience uses, the demand for them and the buying intent behind them. Separate terms that drive business from terms that only drive traffic.
3. Technical SEO audit. Check crawling, indexing, the sitemap, robots.txt, internal link structure, Core Web Vitals and mobile-friendliness. This is where you find what stops Google from reading and understanding the website.
4. On-page analysis. Review on-page SEO: title tags, heading structure, meta descriptions, content depth and how well each page matches the search intent of its target term.
5. Content and relevance review. Judge whether the content answers the question better than the pages currently ranking at the top. Look for gaps (questions you do not answer) and overlaps (pages competing with each other for the same term).
6. Competitor and gap analysis. Compare your visibility with that of your three to five most important competitors. Where do they rank that you do not, and why?
7. AI visibility check. Test how your company is described in ChatGPT, Perplexity and Google's AI Overviews for the most important questions. More on this below.
8. Prioritise by business impact. Sort the findings by expected business impact relative to effort – not by how red the warning was in a tool.
9. Compile an action list. Finish with a concrete, prioritised list with an owner and a time estimate for each item. That is what the audit delivers.

If you want a broader checklist to work through on an ongoing basis, our SEO checklist 2026 is a good complement to the audit itself.

Which tools do you need for an SEO audit?

Free tools get you a long way, but a business-level audit usually needs at least one paid tool for keyword and competitor data. Here is a reasonable toolkit.

Free – the basics everyone should use:

• Google Search Console – shows how Google sees your website: indexing, search queries, clicks and technical errors. Indispensable and free.
• Google Analytics – links traffic to behaviour and conversions.
• PageSpeed Insights – measures Core Web Vitals and performance page by page.
• Bing Webmaster Tools – relevant now that more AI answers draw on Bing's index.

Paid – for depth and competitor data:

• Screaming Frog – crawls the whole website and reveals technical errors at scale (free up to 500 URLs).
• Ahrefs or Semrush – keyword data, backlink profile, competitor analysis and rank tracking.

You do not need to own every tool. What matters is what you do with the data – the tool only provides the raw material.

SEO audit, SEO analysis or SEO report – what is the difference?

In practice the terms overlap. "SEO audit" is the usual term for a structured review of a website's SEO status, and "SEO analysis" often means the same thing. An "SEO report" is usually the document that comes out of an audit – or the automatic export from a tool, which is not the same as an audit.

If you want to draw a distinction, "audit" often leans towards the technical review (the first layer), while "analysis" more often also covers the business and content perspective. But no standard governs the choice of word – always ask what is actually included, whatever word the agency uses.

AI visibility in the SEO audit in 2026

This is the layer most audits skip – and in 2026 it is no longer optional. When a buyer asks ChatGPT, Perplexity or Google's AI Overviews "which agency do you recommend for X", a judgement is made that no traditional SEO report captures. Optimising for it is called AEO, answer engine optimization.

The key insight: AEO is not just about visibility – it is about AI describing your company correctly. If an AI lacks information about you, it tends to answer "no" rather than "I don't know" – and you lose the deal before you have even made contact. So an SEO audit in 2026 must test what AI actually says about you.

Concrete things to check in the AEO layer:

• Are you cited in AI Overviews for your most important questions, and are you described correctly?
• Is the content structured for citability – self-contained paragraphs, clear answers, semantic chunking?
• Do schema markup and structured data support the AI's understanding of what you do?
• What do your brand mentions look like on other websites – a signal AI gives a lot of weight to.

We go deeper into how you actually appear in these answers in AI Overviews and GEO and in the broader AI SEO 2026 guide. This is also why AEO is rarely solved by technology alone: it takes business understanding – knowing what the buyer needs to know – and copy that AI wants to cite, structured so that machines can read it.

How to prioritise the actions by business impact

An audit without prioritisation is just a list. The decisive step is to rank the findings by expected business impact relative to effort – not by how serious a warning looks in a tool. A simple model:

Priority = (traffic potential × conversion rate × margin) ÷ effort

The logic: a fix on a page with buying intent, a reasonable conversion rate and a high margin almost always beats a technical detail on a page no buyer lands on – even if the tool flagged the latter in red. Always divide the business value by the effort, so that quick wins with a big impact end up at the top.

This is also where an automated score falls short and can even mislead: it ranks by technical severity, not by money. As a result, teams often spend weeks chasing a perfect score on pages that make no difference to the business.

Which keywords should you prioritise?

Enter your candidate keywords. The tool combines four factors into a priority score (0–100) – and deliberately gives search volume the lowest weight. A keyword with high volume but low business relevance should not beat a niche keyword with clear purchase intent.

KeywordVolume/monthKD 0–100Search intentInformationalCommercialTransactionalBusiness relevance1 – Low23 – Medium45 – High

82Prioritise

×

KeywordVolume/monthKD 0–100Search intentInformationalCommercialTransactionalBusiness relevance1 – Low23 – Medium45 – High

61Evaluate

×

KeywordVolume/monthKD 0–100Search intentInformationalCommercialTransactionalBusiness relevance1 – Low23 – Medium45 – High

42Low priority

×

+ Add keywordRank by score

• Business relevance30%
• Search intent25%
• Competition (KD)25%
• Search volume20%

How the score is weighted: Business relevance 30% · Search intent 25% · Competition (KD) 25% · Search volume 20%. Volume is counted on a logarithmic scale – double the volume does not give double the score.

An indicative prioritisation, not a definitive answer. Always combine it with market knowledge and with what buyers actually ask Google or AI when they are close to a buying decision.

More free tools

What does an SEO audit cost?

The price varies greatly with scope. An automated free scan costs nothing but only scratches the technical surface. A thorough, manual audit covering all three layers – the technical layer, business context and AI visibility – is consultancy work, and the price depends on the size of the website, the number of keywords and how deep the competitor analysis needs to go.

Rather than quote unreliable standard figures, we suggest looking at how SEO services are priced in general and what you actually get for your money – we cover this in What does an SEO agency cost in 2026?, and the same price–value logic applies to an SEO audit. Always ask what is included, and ask for a concrete deliverable (a prioritised action list) rather than just a score.

Should you do the audit yourself or hire an agency?

You can absolutely do an SEO audit yourself. With Google Search Console, PageSpeed Insights, Claude and a free crawl you can get a long way on the technical layer, and you can handle some of the keyword work on your own. For many smaller websites, that is enough to find the first obvious wins.

The business layer, and the strategy for actually ranking for the right things, are harder to do yourself. Separating keywords that drive revenue from those that only drive traffic, reading the competitor gap correctly and judging how you are described in AI answers takes experience – and that is usually where the value of an audit lies. The combination is what makes the difference: strategy, copy, business understanding and a technical foundation in the same audit.

If you want an audit that covers all three layers – including how you appear in ChatGPT, Perplexity and Google's AI Overviews – get in touch and we will look at where your biggest business wins actually are.