Google Manual Actions: What Businesses Need to Know and How to Recover

A Google manual action is not the same as a normal ranking drop.

It means someone at Google has reviewed your site and decided that part of it, or the whole site, does not comply with Google’s spam policies. Google says a manual action can cause pages or an entire site to rank lower or disappear from search results, often without any obvious warning to users.

The problem is that many businesses treat a manual action like an ordinary SEO task: fix a few URLs, monitor rankings and work through the rest over time.

That is not how Google tells you to handle it.

Google’s own documentation contains three points that are particularly important:

Fixing the issue on just some pages will not earn you a partial return to search results.

Google needs to be able to reach the affected pages.

The page patterns shown in Search Console are not necessarily the full list of affected pages.

Put those three points together and the message is quite simple:

You need to understand the full problem, fix everything that needs fixing, make sure Google can see the result, and only then ask Google to review the site.

There is also an important change to Google’s site reputation policy that UK businesses need to understand.

From 30 August 2026, site reputation manual actions no longer affect search results shown to users in the European Economic Area (EEA). Outside the EEA, the manual action still applies. Google says affected sections may also be separated from the main domain in its systems and ranked independently in the EEA.

The UK is not part of the EEA.

That means a UK business with a site reputation issue cannot assume that the new European treatment protects UK search traffic.

For businesses publishing third-party commercial content, this is now an important difference between markets.

1. What actually happens when you get a manual action?

The first mistake is thinking that the URLs shown in Search Console are a complete list of everything that needs attention.

They are not.

Google says the Manual Actions report shows patterns of affected pages. Those patterns can cover part of a site or the entire site, and Google specifically says that not every page matching the pattern is necessarily affected.

So seeing something like:

example.com/real-estate/*

does not mean:

“These are the only URLs we need to fix.”

It means you have been given an indication of where Google found the problem.

You still need to investigate the site yourself.

That matters more on larger websites.

There is no partial recovery

Google is unusually clear about this.

If you have 10,000 affected pages and fix 5,000, you do not get 50% of the benefit.

Google says fixing only some pages will not give you a partial return to search results. If there are multiple manual actions, they all need to be addressed.

That changes how the project should be managed.

A normal SEO project can often be phased:

• Fix the highest-value pages
• Measure the impact
• Expand the work
• Repeat

A manual action is different.

You need to treat a manual action as one complete recovery project. This is something Konrad Szymaniak, our founder and managing enterprise SEO consultant, specialises in!

This is one reason large organisations can get stuck. A team does most of the work, sees no immediate improvement and assumes the diagnosis was wrong.

It may not have been wrong.

They simply had not finished the job.

2. Do not hide the problem from Google

One of the easiest mistakes I see SEO teams make is also one of the most understandable.

You discover spammy pages, poor content or problematic third-party content, so you try to make those pages disappear:

• Block them in robots.txt
• Add noindex
• Put them behind a login
• Put them behind a paywall
• Remove them from important internal links

The problem is that Google needs to be able to see what you have changed.

Google’s general manual action guidance says affected pages should not require a login, sit behind a paywall, or be blocked by robots.txt or a noindex directive when the pages need to be reviewed. Google recommends using URL Inspection to check that it can access them.

There is an important exception for site reputation abuse.

Google specifically lists noindex as one possible way to deal with third-party content that violates the policy. However, it also says not to block that content with robots.txt, because doing so can prevent Google from seeing the noindex directive.

So the practical rule is:

Do not use robots.txt as a substitute for fixing a manual action.

Sometimes noindex is part of the correct fix. What matters is understanding the specific manual action and making sure Google can verify the result.

3. The pages listed in Search Console are a starting point

This is another area where teams can take Google’s wording too literally.

Google says the Manual Actions report shows patterns of affected pages, not necessarily every affected URL.

That means your own audit matters.

For example, imagine a publisher receives a manual action related to user-generated spam.

Search Console shows a handful of URLs under /forum/.

It would be a mistake to clean those five URLs and stop.

You should look at:

• Forums
• Comments
• User profiles
• Guestbooks
• Upload areas
• Internal search pages
• Other areas where visitors can create or submit content

Google itself tells site owners to identify these areas and look for spammy accounts, commercial usernames, irrelevant links, advertisements, gibberish and automatically generated content.

The question is not:

“Which URLs did Google show us?”

The better question is:

“Where else on this site could the same problem exist?”

4. Manual actions are easy to miss

Manual actions are particularly dangerous because there may be no obvious warning for visitors.

Google says most manual actions result in pages or entire sites being ranked lower or removed from search results without any visible indication to the user.

That is very different from a security issue.

A hacked site may show a browser warning or a warning label in Google Search.

A manual action may simply look like:

“Why has our organic traffic disappeared?”

The notification sits in Search Console.

That means businesses need to know:

• Who has access to Search Console?
• Who receives the notifications?
• Is that person still with the business?
• Is an agency receiving everything?
• Is anyone checking the Manual Actions report regularly?

If the only person receiving those notifications left the company two years ago, your early warning system effectively does not exist.

5. The EEA change is important for UK businesses

On 30 August 2026, Google changed how its site reputation policy operates in the EEA following discussions with the European Commission.

For searches conducted inside the EEA, the manual action itself no longer affects the search results.

Outside the EEA, it still does.

Google also says the affected section of a site may be separated from the main domain and ranked independently over time.

This creates an important market difference.

A publisher operating the same third-party content model could now see different treatment depending on where the searcher is located.

A German user is inside the EEA.

A UK user is not.

So a UK publisher should not assume that the European change removes its exposure to site reputation manual actions in the UK.

There is another important point here.

The EEA change does not mean that Google has decided this type of content is acceptable.

Google says it still considers the practice a problem. Its current documentation says that, in the EEA, affected sections may be treated as separate from the main site and ranked on their own merits.

So the risk has changed. The underlying policy has not simply disappeared.

6. What types of businesses are most exposed?

Publishers and media companies

This is one of the clearest areas of risk.

Examples can include:

• Commercial content produced by third parties
• Coupon and deals sections
• Betting content
• Product reviews
• “Best” or comparison pages
• Affiliate content
• Content created mainly to take advantage of the authority of an established publisher

Google’s site reputation policy is specifically concerned with third-party content being placed on a trusted site mainly because the host site’s existing ranking signals can help that content rank.

That does not mean all third-party content is a problem.

Google explicitly says third-party content by itself is not a violation. The context, level of editorial involvement and purpose of the content matter.

Forums, communities and user-generated platforms

If users can create accounts, publish posts, upload files or create profiles, you have a potential spam surface.

Google specifically mentions forums, guestbooks, social platforms, file uploaders, free hosting services and internal search pages in its manual action guidance.

Sites with indexable internal search pages

Internal search can become a surprisingly large source of spam.

Google specifically warns about internal search results where queries appear to be promoting unrelated third-party websites or services.

Ecommerce sites and large websites using structured data

Structured data can also lead to manual actions.

Google’s current documentation includes issues such as:

• Markup that does not match the visible page
• Markup for content that does not exist on the page
• Aggregating multiple items into one structured data element
• Job listings using misleading information
• Job listings with incorrect or missing validThrough
• Structured data on the wrong type of page

The important lesson is that passing a structured data validator is not enough.

The markup also needs to accurately describe what users can actually see on the page.

Job boards and recruitment websites

Job boards have a particularly large structured data surface.

Google lists a number of manual action issues specifically related to JobPosting, including misleading locations, paid applications, expired jobs and job markup on listing pages.

News and Discover publishers

News and Discover sites have additional policy requirements.

For some manual actions, Google asks for evidence of changed editorial practices, new editorial policies and an editorial board with a history of improved practices.

That is not simply a technical fix.

It may require a genuine change to how the organisation publishes and reviews content.

Hosting companies and SaaS platforms

If customers can create websites or pages on your platform, their behaviour can create risk for the platform itself.

Google says it may take action against an entire free hosting service when a significant proportion of the sites hosted there are spammy.

Businesses buying or acquiring websites

A manual action can survive a change of ownership.

Google explicitly has guidance for businesses that have bought a site which violated its spam policies before the acquisition. The advice is to fix the issues and explain in the reconsideration request that the site has been acquired and now complies.

That makes the Manual Actions report a sensible part of website due diligence.

A five-minute Search Console check can uncover a problem that becomes a much larger project after completion.

7. What should businesses actually do?

The biggest mistake is starting to make changes before understanding the full problem.

Start with the facts.

7.1 Check Search Console first

Confirm:

• What manual action has been applied?
• Which pages or patterns are listed?
• Is the whole site affected?
• Are there multiple manual actions?
• Who receives the notifications?
• When was the action first reported?

Read the full description rather than just the headline.

Then investigate the site independently.

Do not assume the examples in Search Console represent the entire problem.

7.2 Audit the whole area where the problem could exist

If the issue involves user-generated content, investigate every place users can contribute.

If it involves structured data, audit templates and page types across the site.

If it involves third-party content, look beyond the URLs listed in Search Console.

If it involves hacked or injected content, check newly created URLs and unexpected traffic patterns.

For user-generated spam, Google itself recommends using the site: operator with unrelated commercial or adult keywords to find unexpected content. It also recommends checking server logs for sudden traffic spikes and strange URL patterns.

For example:

site:example.com viagra
site:example.com "free movie watch online"
site:example.com casino
site:example.com "cheap replica"

You are looking for things such as:

• Completely unrelated pages
• Gibberish
• Auto-generated text
• Commercial usernames
• Unrelated outbound links
• Suspicious internal search results
• Unexpected new URLs

7.3 Make the decision about problematic content

For site reputation issues, Google currently gives several possible options:

• Move the content to a new domain
• Use noindex
• Rework it into genuine first-party content
• Remove it

But Google also warns that simply moving problematic content to a subdomain or subdirectory may be seen as an attempt to get around the policy. Moving it to another established site can also move the problem rather than solve it.

In other words, there is no clever URL structure that magically solves the problem.

You need to decide whether the content genuinely belongs on the site and whether the business is prepared to bring it properly under its own editorial control.

7.4 Make sure Google can see the changes

Before submitting a reconsideration request, check the pages yourself.

Use URL Inspection.

Check:

• HTTP status
• Crawlability
• noindex
• X-Robots-Tag
• robots.txt
• Redirects
• Rendered HTML
• The actual content Google can see

For large sites, this should be automated where possible rather than checked manually on a handful of URLs.

For example:

for (const url of affectedUrls) {
  const response = await fetch(url, { redirect: "manual" });

  if (response.status !== 200) {
    throw new Error(`${url} returned ${response.status}`);
  }

  const html = await response.text();

  if (/noindex/i.test(html)) {
    throw new Error(`${url} contains a noindex directive`);
  }
}

The exact implementation will depend on your stack, but the principle is simple:

Do not rely on five spot checks when thousands of URLs are involved.

8. Fix the problem at its source

Removing the pages is only half the job if the system can simply create more of them tomorrow.

For user-generated spam, Google’s guidance includes several practical measures:

• Moderate new accounts and submissions
• Block obvious spam terms
• Monitor forums, profiles and comments
• Use rel="ugc" for appropriate user-generated links
• Monitor internal search pages
• Check file upload areas
• Keep interactive content organised so it is easier to monitor

The key question is:

Why did this happen in the first place?

If a spammer created 20,000 pages because your CMS allowed anyone to publish indexed content immediately, deleting the 20,000 pages is not enough.

You also need to fix the weakness that allowed 20,000 pages to be created.

9. Structured data needs a proper audit too

A common mistake is treating structured data as a purely technical validation exercise.

It is not.

The markup should describe the content users can actually see.

Check for:

• Markup that does not match visible content
• Markup for content that is not present
• Reviews that do not meet Google’s requirements
• Product markup on something that is not a product
• Incorrect job locations
• Expired jobs without the correct validThrough
• Individual item markup being used incorrectly on listing pages
• Multiple items being incorrectly combined into one structured data element

This is also why structured data checks should form part of your development process rather than being left until an SEO audit.

10. Check for cloaking, hidden content and redirects

Large websites can develop problems without anyone deliberately trying to manipulate Google.

Third-party scripts can introduce redirects.

CMS changes can expose hidden content.

Plugins can add unwanted code.

Developers can accidentally create different experiences for Google and users.

Google’s manual action documentation specifically covers cloaking, hidden text, sneaky redirects and sneaky mobile redirects.

For a serious investigation, compare:

• What a normal user sees
• What Google sees
• What URL Inspection reports
• What happens on mobile
• What happens when different referrers or user agents are used

Third-party scripts deserve particular attention when investigating mobile redirects.

11. Write the reconsideration request properly

The reconsideration request should not be:

“We fixed the issue. Please review.”

Google tells you exactly what a good request should contain:

1. What was wrong
2. What you changed
3. What happened as a result

That sounds obvious, but many requests focus almost entirely on the second point.

A much stronger request explains the problem clearly.

For example:

We identified 14,287 pages containing third-party content that did not meet Google’s site reputation policy.

We removed 9,841 pages, rewrote 3,926 as first-party content and added noindex to 520 pages that are not intended for Search.

We also changed the publishing process so third-party commercial content now requires editorial approval, author attribution and review before publication.

That gives the reviewer something concrete to assess.

Numbers, examples, dates and evidence are far more useful than vague assurances.

12. Some fixes are organisational, not technical

This is particularly important for publishers.

If Google asks for evidence of changed editorial practices, new policies and an editorial board with a history of improved practices, changing a few URLs is not enough.

That means the person responsible may not be the SEO manager.

It could involve:

• Editorial leadership
• Content teams
• Legal or compliance
• Product
• Engineering
• Marketing
• Senior management

A manual action can therefore become a business process issue, not just an SEO issue.

13. The recovery process

The process should be simple:

1. Check Search Console

Read the full manual action and confirm who receives notifications.

2. Understand the real extent of the problem

Use the patterns in Search Console as a starting point, not the complete scope.

3. Check for multiple manual actions

All of them need to be addressed.

4. Audit the site

Find every area where the same problem could exist.

5. Fix the underlying cause

Do not simply remove the visible symptoms.

6. Make sure Google can access the affected pages

Use automated checks where possible, then confirm with URL Inspection.

7. Add safeguards

Stop the same issue from being recreated.

8. Document what changed

Record URLs, numbers, dates, examples and process changes.

9. Submit the reconsideration request

Explain the problem, the fix and the evidence.

10. Do not keep resubmitting

Google says reconsideration reviews can take several days or weeks, and sometimes longer for link-related cases. It also says not to submit another request while the existing review is still outstanding.

11. Keep monitoring the site

Do not stop cleaning up problems just because the request has been submitted. Google recommends continuing to monitor and remove spam while a review is in progress.

14. Prevention is much cheaper than recovery

A manual action should not be something a company checks only after traffic collapses.

Add it to your regular SEO monitoring

Have a named person check:

• Manual Actions
• Search Console messages
• Security Issues
• New unexpected URLs
• Large traffic changes
• Spam appearing in user-generated areas

Add it to website acquisitions

Before buying a domain, check:

• Manual Actions
• Search Console history
• Outstanding reconsideration requests
• Major indexing problems
• Unusual URL patterns
• Previous spam issues

Have a clear policy for third-party content

If a commercial partner wants to publish content on your domain, ask:

• Why is it being published?
• Who created it?
• Who edits it?
• Who is responsible for it?
• Is it genuinely useful to your audience?
• Is it integrated into your site?
• Is the commercial relationship clearly disclosed?

This matters even more for UK publishers now that the EEA treatment is different.

Review every new interactive feature

A new:

• Comment system
• Forum
• Profile area
• Upload tool
• Search feature
• Community section

can create a new spam surface.

SEO review should happen before the feature goes live, not after Google finds the problem.

15. What we are watching next

Different treatment by market

The August 2026 site reputation change shows that Google Search policies can now have different effects depending on where the searcher is located.

That matters for international businesses.

A website can no longer assume that one SEO risk profile automatically applies in exactly the same way everywhere.

The UK’s position

The UK sits outside the EEA, so businesses targeting UK searchers should not assume that the EEA carve-out protects them.

That makes third-party commercial content a question UK publishers should actively review rather than wait for a manual action.

More scrutiny of large-scale content

Google’s spam policies already include scaled content abuse, including large volumes of low-value content created using generative AI or other automated methods.

The issue is not simply whether AI was used.

The bigger question is whether a business is publishing large amounts of content without enough original value, editorial oversight or quality control.

Manual actions becoming part of due diligence

For publishers, ecommerce businesses, marketplaces and acquisitions, checking Search Console should become as normal as checking technical infrastructure, security and analytics.

Because finding the problem before an acquisition is considerably easier than inheriting it afterwards.

About Szymaniak Digital

Szymaniak Digital is an enterprise AI SEO consultancy working with senior marketing, SEO, content and technical teams.

We work on complex SEO problems where the answer is rarely just “change the title tag”.

That includes manual action investigations, site reputation issues, large-scale content problems, technical SEO, migrations and the systems needed to prevent the same problems from coming back.

For businesses publishing third-party commercial content on a UK domain, the August 2026 Google update makes the issue particularly relevant.

The EEA now has a different treatment for site reputation manual actions.

The UK does not fall within that carve-out.

That is something UK publishers and marketing teams should understand before they have to deal with a manual action.

Contact Us!