Free SEO Audit: What It Finds and What It Never Will

"Free" in SEO auditing means two very different things in practice. Either you get a thirty-page PDF full of charts and a recommendation to "optimise your content", or you get a list of specific URLs with specific errors you can verify on your own site in a browser tab. The gap between those two is huge, so I want to lay out what can genuinely be detected from publicly available pages, what cannot, and where the line sits between "I found a problem" and "I have the fix written".

What a free SEO audit does find

The crawl goes through your public pages the same way a search engine does. That defines the boundary of what it can see: anything in the HTML, in the response headers, in the sitemap, and in page load measurements.

Titles and meta descriptions

The most common finding on shops with dozens of categories. Duplicate titles appear on their own — filtering, pagination, product variants, categories sharing one template. The audit shows which pages carry the same title, which ones run too long and get cut off in the results, which have no title at all, and which have a description that just repeats the category name. Same for missing or generic descriptions, where Google writes its own snippet from the body text — usually worse than what you would have written.

Heading structure

Pages with no H1, pages with five H1s, skipped heading levels, H2 used purely as a design element. It is not fatal, but on long pages and articles it changes how a search engine understands what the page is about.

Image alt text

On ecommerce sites this means hundreds of images with no alt text, or with alt text like IMG_20240113.jpg. It affects image search and accessibility.

Structured data

Broken or incomplete schema markup is a quiet problem. Nothing shows up as an error on the page — the stars, price, availability or breadcrumbs simply never appear in the search results. The audit shows where markup is missing, where it is invalid, and where a required field sits empty.

Page speed

Which specific templates load slowly and why — uncompressed images, blocking scripts, no caching headers. What matters here is knowing which page types are slow, not an average across the whole site.

Indexability and duplication

Pages excluded from the index that belong in it. Wrong canonical tags. Redirect chains. Internal links pointing to 404s. A sitemap listing URLs that no longer exist.

All of this can be established from the outside, with no admin access and nothing deployed. On my side the output is ready in roughly three minutes from the moment you enter the domain.

What a free audit will not tell you

This is the unpopular part, and it matters more than the feature list.

Backlink quality. Who links to you, whether those links look natural, whether someone is dragging you down with spam — none of that is readable from your HTML. Different discipline, different data.

Actual rankings and the queries you get found for. Without Google Search Console connected, any position estimate is just an estimate. That is why my article suggestions run on real Search Console data — queries where you already get impressions but have weak CTR or sit on page two. Without that access, any tool can only offer you a topic "by search volume", which is a much weaker starting point.

Anything behind a login. Cart, checkout flow, customer account, drafts of unpublished pages. The crawl follows public pages, so it does not see them.

Business context. The tool does not know that the category with the worst titles happens to be your highest-margin one, and that another is being discontinued. Always sanity-check the fix order against what actually brings in revenue.

Deep competitive content analysis. Whether your product copy beats a competitor's is a judgement call, not a measurement.

Any guarantee of results. I do not promise that, and nobody sensible does. Rankings move slowly — from deploying a fix to a visible change in Search Console you are looking at weeks, not days. Anyone promising you the top spot within a few days is lying.

A list of errors versus a finished fix

Most free audits stop at the diagnosis. You get "title is too long, shorten it" or "add a meta description". That is like a garage telling you your brake pads are worn and sending you home.

Finding the error is not the hard part. The hard part is writing 180 titles so that each one carries the right keyword, is not a duplicate of the others, fits the width of a search result, and still reads like a sentence a human wrote. That is days of work.

So for the findings that matter, I write the fix text itself. Not advice — a finished title, a finished meta description, a finished alt text, completed structured data. You review it, approve it or rewrite it. Deployment goes through the WordPress plugin, through the API, or as a CSV export if you run your own system. Nothing gets overwritten without your approval.

After deployment I track what happened: 28 days before against 28 days after, at individual page level. That way you know which change worked and which was wasted effort — instead of a vague sense that "we optimised something".

What order to fix things in

With the audit in front of you, the reflex is to start at the top of the list. Better to start by impact:

1. Errors blocking indexation. Wrong noindex, misconfigured canonicals, important URLs missing from the sitemap. Until a page is in the index, its title is irrelevant.
2. Titles and descriptions on your highest-traffic, highest-margin pages. The effect shows up fastest here, because those strings directly decide whether anyone clicks.
3. Duplicate category titles. The classic ecommerce finding, and typically the biggest volume of work, so it pays to handle it in batches.
4. Structured data on products. It changes how prominent your listing looks in the results.
5. Speed and alt text. Important, but rarely the most painful thing to tackle first.

If you get through the first three, you have covered most of what a typical agency bills for in month one. Where the sensible boundary sits between doing it yourself and paying someone, I cover separately.

When a free audit is enough

It is enough when you need to establish whether you have a technical problem at all and how big it is. For that it is ideal: within minutes you see a list of specific pages with specific errors, and you can decide whether to act.

It is not enough when your organic traffic has been declining for months with no obvious technical cause. In that case the problem usually sits in the content, in the links, or in the fact that competitors are simply doing more — and none of that gets solved by rewriting meta descriptions.

FAQ

Do I need to give access to my site for the audit to run? No. I crawl public pages, so the domain is enough. You will need access later, at deployment — or a Search Console connection if you want article topics based on real queries.

Will an automatic fix ruin the titles I wrote myself? No, because nothing is deployed without approval. You see every fix before it goes live and can reject or rewrite it.

How long until I see a change in Google? It depends how often the search engine crawls you. On smaller sites it tends to be weeks. That is why I measure on a 28-day before-and-after window rather than daily fluctuation.

Is an audit worth it on a new site with no traffic? Yes, precisely for the technical part. It is far easier to have titles, descriptions and structured data correct from the start than to repair them across a thousand URLs later.

What if I run a custom system, not WordPress? You get the fixes through the API or as a CSV export with URLs and new values. You import them the way you normally import anything into your system.

What to do now

Enter your domain and let the crawl run. Within minutes you see how many findings are waiting and exactly where they sit. Then you decide whether to work through them yourself or have the fix text written and just click approve.

#seo obsah