How to Do an SEO Audit Without Buying a Tool

Google's free tools answer most of the questions an audit asks, and their published limits are where the genuinely useful findings live. That second half is what separates this from the usual free-audit article, which lists tools and skips what each one refuses to tell you — including two that Google shut down years ago and guides still recommend.

This is the sequence we actually run before quoting work, in the order we run it, with the limit attached to each step. If you want the basic checks in a few seconds first, our free SEO checker runs on the same logic.

1. Can Google reach the site at all?

Start with the robots.txt report in Search Console. It shows the robots.txt files Google found for the top 20 hosts on your site, when each was last fetched, and any warnings or errors — and it lets you request a re-fetch when something urgent has been blocked.

Read the file yourself as well, because the error that matters is usually not a syntax error. A Disallow rule that was correct on staging and shipped to production is valid robots.txt and a catastrophe. So is a named user-agent group added below a wildcard group, which replaces the wildcard rules for that crawler rather than adding to them.

What it cannot tell you: whether blocking was intended. Only you know that, which is why this step is a conversation and not a check.

2. What does Google actually have?

Open the Page Indexing report and read the "not indexed" reasons before the indexed count. The reasons are the audit. "Discovered — currently not indexed" says Google knows about the URL and has chosen not to spend a crawl on it. "Crawled — currently not indexed" says it fetched the page and decided against it. "Duplicate without user-selected canonical" says your templates are producing pages Google cannot tell apart.

Those three cover most of what a paid crawl would surface, and they come from Google's own index rather than a third party's model of it.

What it cannot tell you: why. The report names the state, not the cause, and the cause is usually visible only by opening a few of the URLs listed under each reason.

Infrastructure checks such as redirect chains, www consistency and the status code a missing page returns, measured on 200 popular Turkish sites, are in the technical SEO checklist.

3. What does Google see on one page?

Now use the URL Inspection tool on a handful of representative URLs — not on everything, because Google states there is "a daily limit of inspection requests for each property that you own" and separately "a per-property daily limit of live inspections," without publishing either number.

Two readings matter and they are different things. The indexed result is what Google last crawled and is currently using. The live test fetches the page now, and Google notes that "the information shown in the live test can differ from the indexed URL" — which is exactly what you want when confirming a fix.

Then the sentence that should be printed above every audit: "URL is on Google" doesn't guarantee search visibility — only eligibility. The tool reports whether a page can appear, not whether it does. A perfectly indexed page that gets no traffic has a demand or content problem, and no amount of technical work will move it. Our SEO guide covers that side of the diagnosis.

Google also lists what this tool does not test at all: quality and security guidelines, manual actions, security issues, and content removals. Those live elsewhere in Search Console, and an audit that never opens the Manual Actions page has skipped the one finding that would explain everything.

4. Does the page render without JavaScript doing the work?

In the live test, look at the rendered HTML and the screenshot rather than the verdict. This is where most real problems in modern sites are found: content that exists only after client-side fetching, navigation that never appears in the HTML, and main content replaced by a loading state at the moment Google looks.

This is also the step that has quietly changed. Google's own notice reads: "Update on December 1, 2023: The Mobile Friendly Test and the Mobile Usability report have been retired." Any guide still sending you to those tools was written before that date, or copied from something that was — a useful test of whether the guide you are reading has been maintained.

5. Is it fast for real users, not in a lab?

Run the page through PageSpeed Insights. The important distinction is on the same screen: field data comes from the Chrome User Experience Report and describes what actual visitors experienced, while the Lighthouse score below it is a simulation run on Google's hardware with synthetic throttling.

Treat the field data as the measurement and the lab score as the diagnosis. A site can score 62 in the lab and pass Core Web Vitals in the field, or score 98 and fail, and we have seen both. Sites without enough traffic get no field data at all, which is a limit rather than a pass. We went through how to act on each metric in our Core Web Vitals guide.

6. Is the structured data valid, and is it for something real?

The Rich Results Test validates the markup on any URL, including ones you do not own — which makes it the tool for checking competitors as well as yourself.

The finding worth looking for is not invalid markup. It is valid markup for a result type that does not exist for your kind of page, which costs implementation time and returns nothing. We wrote about which types are worth the effort in our schema markup guide.

7. The half no tool does

Everything above is the machine-checkable half. The rest of an audit is reading the site as the person searching would, and no crawler substitutes for it:

• Does each page answer the question its title promises, in the first screen? (The on-page elements that still matter covers what Google documents here, and what it does not.)
• Are there several pages competing for the same query, so Google has to choose and keeps choosing differently?
• Does anything on the site say who wrote it and why they would know?
• If you removed the branding, would the pages be distinguishable from a competitor's?

That list finds more revenue than steps one to six in most audits we run. It is also the part that cannot be outsourced to a subscription, which is why it is missing from articles written to sell one.

When a paid crawler earns its price

We are not against tools and we use them. They are worth it when the problem is scale rather than understanding: thousands of URLs, a migration to verify, faceted navigation generating combinations nobody planned, or an internal-link structure too large to hold in your head. A crawler finds the pattern in an afternoon that would take a week by hand.

Below a few hundred pages, the free tools and a spreadsheet cover it, and the budget is better spent fixing what you find. If a proposal's first line item is a tool subscription for a 40-page site, that is worth asking about — along with the other questions we would ask any agency.

The order, as a checklist

1. robots.txt report — can Google reach it, and was any blocking intended?
2. Page Indexing report — read the not-indexed reasons before the indexed count.
3. URL Inspection — a few representative URLs, indexed version first, then live.
4. Manual Actions and Security Issues — the pages that explain everything when they are not empty.
5. Rendered HTML in the live test — does the content exist without client-side fetching?
6. PageSpeed Insights — field data is the measurement, the lab score is the diagnosis.
7. Rich Results Test — valid, and for a result type that exists.
8. Read the pages — the half that finds the money.

If you would rather have this run by someone who has done it a few hundred times, send us the domain and we will report what we find against these steps — or see what a full engagement contains on our SEO services page.