An SEO audit tool that shows what to fix next

What an agency does, and what this site audit tool does

Three steps from crawl to follow-up

1. Tell us your site

   Tell us which site to audit when you ask for an invitation. We set it up for you.

2. See what needs work

   Seoverk crawls your HTML pages and gives you a list of what to fix, with the most important issues at the top.

3. Fix and check again

   Make the changes yourself, ask your developer or use your own AI agent. The next crawl shows which issues are fixed.

What Seoverk checks

Issues have high, medium or low severity and are marked as findings or advice. These are examples from the live beta.

• Crawling and indexing

  Finds noindex pages, canonicals pointing to dead URLs, hreflang errors and missing robots.txt files.

• Status codes and redirects

  Flags internal links to 4xx and 5xx pages, broken external links and redirect chains with several hops.

• Titles and headings

  Shows missing or duplicate titles and meta descriptions, missing H1 headings and skipped heading levels.

• Content and internal links

  Finds placeholder text, near-duplicate pages, deep pages and pages with few inbound internal links.

• Structured data and images

  Checks invalid JSON-LD, product markup errors, broken images and missing alt text.

• Sitemaps and mobile

  Finds redirecting or noindex URLs in a sitemap and pages without a viewport tag.

• Technical setup and response

  Checks mixed content, invalid TLS certificates, missing security headers and slow HTML responses from the server.

• AI search and GEO

  Checks AI crawler access in robots.txt, llms.txt and page structure signals that make content harder to cite.

Connect your own AI agent with SEO MCP

Seoverk is an SEO crawler with an MCP server at https://app.seoverk.com/mcp. Your AI SEO agent connects with a key in an HTTP header. It can read crawls, start a crawl within the beta limits and re-crawl a single page.

Claude Code

claude mcp add --transport http seoverk https://app.seoverk.com/mcp --header "X-API-Key: $SEOVERK_KEY"

Another MCP client over HTTP

{
  "mcpServers": {
    "seoverk": {
      "type": "http",
      "url": "https://app.seoverk.com/mcp",
      "headers": {
        "Authorization": "Bearer ${SEOVERK_KEY}"
      }
    }
  }
}

Replace ${SEOVERK_KEY} with your key if your client does not load it from an environment variable.

We create your key when you receive an invitation. A page to manage your own keys will come later.

Planned additions

• Coming soon

  A task list for agents

  Your agent will be able to take a task, change your site and use the next crawl to confirm the fix.

• Coming soon

  White label SEO audit tool

  Agencies will be able to put their own brand on reports and sell the analyses under that brand.

• Coming soon

  Seven-day self-service trial

  You will be able to create an account and try limited access for 7 days.

SEO audit tool for agencies and their clients

Each client has a separate workspace and key today. A key reaches that client's crawls and cannot reach another client's data. Use the same list of what to fix when you report progress to the client.

An agency key that reaches all your clients is planned for later. Branded reports are also planned for later.

Beta pricing and limits

How to do an SEO audit, step by step

You can audit a website without buying a monthly service. Start with the pages you want people to find in search. Record each issue, the affected URL and the change you make, so you can check the outcome later. A website audit tool saves time on a large site, but you can follow these steps by hand on a small one. Keep a dated copy of your results before you edit anything. That baseline matters when a later crawl shows a different set of pages.

1. Crawl the site

Start at the home page and follow internal links. Record which URLs are reachable, which return errors and which important pages do not appear in the crawl. A crawler can collect status codes, titles, descriptions and links in one report. Seoverk reads HTML and covers up to 500 URLs per crawl during the beta, so larger sites need another way to cover the rest. Compare the result with your sitemap. A page listed in the sitemap but absent from internal links may be hard for visitors to discover. Keep the crawl date with the report so you can compare it with later work.

2. Check whether pages can be indexed

Open robots.txt and look for rules that block areas you want indexed. Check important pages for a noindex instruction. Then inspect their canonical URLs: each should point to the intended, working page. A canonical that points to a dead URL, or a loop between pages, can confuse the choice of preferred page. Read the sitemap and remove URLs that redirect, fail or carry noindex. Make sure robots.txt points to the correct sitemap as well. Review any broad rule before publishing it. One mistaken directive can hide a whole section from crawlers.

3. Fix error responses and broken links

Find internal links to 404 and 5xx pages. If the destination still exists elsewhere, update the link to its current URL. A permanent move may also need a redirect, but internal navigation should point directly to the destination. Follow redirect chains and shorten those that pass through several URLs. Check external links that fail, then decide whether the source has moved or the link should be removed. Begin with site navigation and high-value pages, where one broken path can affect many visits. Recheck the target URL after each change instead of assuming the new link works.

4. Review titles, descriptions and headings

Give each important page a title that describes its specific content. Address missing and duplicate titles before fine-tuning length. Write a meta description that helps a searcher decide whether the page answers their need; search engines may display different text. Check that the H1 identifies the page and that subheadings follow a readable order. Two pages with different purposes should not share the same title and heading. Read the body copy too. Placeholder text and near-duplicate pages usually need an editorial change to the page itself.

5. Map internal links

Look at the number of clicks from the home page to your important content. A deep page can be hard to find even when it appears in the sitemap. Add links from relevant articles, categories or navigation pages where they help a reader move to the next useful page. Review pages with one inbound internal link and pages with none. Write link text that tells readers where they will go. Do not move links around simply to improve a score. Check whether the new route through your site makes sense to a person using it.

6. Check structured data and images

Validate JSON-LD and compare each field with what a visitor can see on the page. A product page with inaccurate fields can have unusable product markup. Check whether the home page has appropriate Organization and WebSite markup. Open the list of broken images and correct their paths. Add useful alt text to images that convey information, and set image dimensions to keep the page layout steady. Decorative images do not need a sales description. A technical SEO audit should point you to a specific defect and the URLs it affects.

7. Read Google's own reports

Open Google Search Console, Google's free tool for a site you control. Compare its indexing report with your list of pages that should appear in search. Inspect important URLs that Google has not indexed and read the stated reason before changing anything. Check the submitted sitemap there too. An SEO crawler reports what it can fetch from your site; Search Console reports what Google says it has done with those pages. The two views may reflect different points in time. Check the page as it exists now and the date of the report before drawing a conclusion.

8. Prioritise, fix and repeat

Put each issue in a list with its URL, severity and proposed change. Start with problems that block access or indexing on important pages. Next, handle broken internal links and faults in shared templates that affect many URLs. Minor edits to one page can wait. Make the changes yourself, ask a developer or use your own agent. Run another crawl and compare it with the saved baseline. Mark which issues are fixed, which remain and which are new. Finally, open your most important pages manually to confirm that the visible experience still works.

Frequently asked questions

What is an SEO audit?

An SEO audit checks whether search engines can reach and understand your website. It covers indexing, links, page titles and technical faults. A useful result is a list of what to fix, with important issues at the top.

How much does SEO cost?

The cited agency survey places common monthly fees at USD 500–1,000. An agency may write content, build links, plan strategy and make the changes. Seoverk is free during the beta, with a preliminary planned price of USD 50 a month afterwards.

How much does an SEO audit cost?

Agencies price one-off audits very differently, and many offer a basic one free to win clients. Ask what the report covers and who will make the fixes afterwards.

Can I do SEO myself?

Yes. Start with indexability, broken links and clear titles on your important pages. Follow the guide above, then check the pages again after each fix. Bring in a developer when the change needs code you cannot safely edit.

What is a technical SEO audit?

A technical SEO audit checks whether crawlers can reach, process and understand your pages. Typical checks cover status codes, robots.txt, canonicals, sitemaps and internal links. It should identify the affected URLs so you can fix the cause.

Who is behind Seoverk?

Uygar Duzgun AB, a Swedish company. The app interface is mostly in English.

Where is the data stored?

The crawlers and accounts run on a server in Sweden. All traffic passes through Cloudflare, which also serves this site. Read the privacy policy (in Swedish).

Does it replace Screaming Frog?

No. Seoverk has no desktop app: its crawler runs on our server and your agent reaches it through a URL. The focus is different. Screaming Frog has its own MCP in its desktop app.

Can my agent start crawls?

Yes, within the beta limits. It can also re-crawl a single page.

How do permissions work for clients?

Each client has a separate workspace and a key that reaches that client alone. An agency key that reaches all of an agency's clients will come later.

How do I get an invitation?

Email us with your site and your role. We create your workspace and key when we invite you.