Hi, I’m Alex. Today I’ll explain why SEO is completely broken — and how that ties into black hat SEO.
Black Hat SEO (aka “black SEO”) is a bunch of SEO tactics that go against Google’s Webmaster Guidelines and/or Google’s (or any other search engine’s) terms of service.
In Google documentation, a lot of it is covered inside Spam Policies:
https://developers.google.com/search/docs/essentials/spam-policies
Cloaking
Doorways
Expired domain abuse (drop abuse)
Hacked content (“pies”)
Hidden text and links
Keyword stuffing
Link spam
Bot-generated traffic
Malware
Deceptive functionality
Scaled content abuse
Scraped (stolen) content
Sneaky redirects
Site reputation abuse (parasite SEO)
Thin affiliate pages
UGC spam (User-Generated Content spam)
This also usually includes:
PBNs (satellites)
Buying links
Structured data (schema) abuse
Auto-updating the “published” date
Site-level exploits / abuses
Below I go through all of them, one by one.
Quick rundown of each Black Hat SEO category
Here I’ll briefly explain what each black hat method is about. For the most interesting ones, I’ll link deeper articles.
1. Cloaking
Cloaking is when you mask one piece/version of content with another.
Examples:
Googlebot gets a kind of “SEO-optimized” page, while the user sees a clean, readable version (or anything a webmaster wants).
Googlebot is served one version of the page, and the user gets a different one. For example: an article titled “How microloans impact your credit history” shows the original content to Googlebot, but users see something else entirely (like a “best microloan companies” list, etc.).
Users from one country see one version of content, while other countries see a different one. You’ll often see this in the gambling niche: users getting relevant offers depending on their country.
Most of the time cloaking is used to trick the search engine: the SE gets one page version, the visitor gets another.
But it’s not a strict rule. Cloaking can also be used to mislead competitors, or to target very specific — sometimes even single — IPs.
2. Doorways
Doorways are pages or whole sites that funnel the user to the final destination (another site/page).
Doorways are usually “optimized” for low- and mid-volume queries (LV/MV) to get into top SERPs. The user clicks from search to a doorway, and then gets redirected to the money page.
“Optimized” is a big word here. A doorway can be a restored expired domain, stolen content, or else.
Doorway example: a blog post comparing microloans that redirects the user to a specific financial product.
3. Expired domain abuse (drop domain abuse)
One of the most popular and everyone’s favorite SEO methods is working with expired domains (drops).
A dropped domain is a domain/site that the owner “dropped” (let it expire).
Typical expired domain abuse flow:
The owner doesn’t renew the domain (business closed, forgot to renew, etc.)
The domain hits auctions
A new owner buys it
The “abuse” part is that the site already has historical signals in the eyes of the SE: backlink profile, age, topical history, and so on.
From there, the new owner has a bunch of options:
Restore the content from the Wayback Machine — often it ranks pretty well (good for PBNs or later migrations)
Publish his own content on it
Move another site onto it
“Glue” it into another site (redirect, canonical, etc.)
There are more options in practice — I cover drops in details in a separate article.
Drops are valuable because buying one is often cheaper and faster than building a new site from scratch and “powering it up.” And the sweetest type of drop is an EMD that matches the new owner’s goals.
Drop pricing usually starts at around $10 and goes to infinity (20k, 40k+, depending mostly on the drop’s link profile strength).
4. Hacked sites (“pies”)
Pie = a hacked site.
Without the owner’s knowledge, links get placed to your targets — how exactly, read in the next section 🙂. Links can be anywhere, but the classic spots are: sitewide footer links, single links inside content, or even a dedicated page created by the hacker for a specific intent.
Most often hacked sites are sold; less often hackers use them for their own projects.
The most common reasons: weak passwords (bruteforced) and outdated CMS/plugins with known vulnerabilities.
5. Hidden text or links
Hidden text/links is content that users don’t see, but bots do (and it’s visible in the source code).
Usually used in:
Cloaking
Hacked sites
Common hiding methods:
White text on white background
Transparent text
Zero font size
JavaScript tricks (display: none, etc.)
CSS absolute positioning (off-screen)
Text placed behind an image
There are more methods — it’s limited only by the creativity of the webmaster/dev :)
6. Keyword stuffing
Keyword stuffing has existed as long as SEO has — 20+ years (and back in the day it was one of the most effective tactics… thankfully or not, that era is gone).
Stuffing / over-saturation is an excessive amount of keywords matching the intent of the page you’re pushing (or not pushing, hehe). It's often called “SEO text”, although you can stuff keywords not only in the main body text.
Example of stuffed text:
“In our sports store you can buy any sports goods. We offer a wide range of sports products useful for sports lovers, kids in school sports classes, for participating in sports events, and helpful for any of your sports activities or professional sports.”
As of now (2026), keyword stuffing in its literal “overstuffing” form isn’t an effective tactic anymore (in Google for sure). And it can be penalized: either you drop out of the index/serps, or you just lose rankings for a page (or sometimes the whole site, but that’s rarer).
7. Link spam
Link spam is any link created mainly to move rankings instead of helping a reader. It’s the oldest category and still the biggest one, because links are still how a search engine measures trust.
Google’s link spam policy lists, among others:
Buying, selling or trading links (or goods and services in exchange for links)
Excessive link exchanges and “partner” pages created only for linking
Automated link creation
Paid articles and advertorials with optimized anchor text
Low-quality directory links, widely distributed footer or widget links, and optimized links in forum signatures
Paid links themselves aren’t forbidden: if a link is marked with rel="sponsored" (or nofollow), Google treats it as advertising. The violation starts when money buys ranking credit.
Since the December 2022 link spam update, Google says its SpamBrain system neutralizes unnatural links and can also spot sites that buy links and sites that exist mainly to pass links on. So a lot of this spend simply does nothing. When it does get noticed, the result can be a manual action: in the Manual actions report it’s called “Unnatural links to your site” (or “from your site” if you were the one selling them). If you inherited bad links, I explain what to do in my article about the disavow tool.
8. Bot-generated traffic
This includes a bunch of bot traffic flavors. I’ll quickly run through the popular/obvious ones:
Automated queries to Google. This is what Google’s policy actually calls “machine-generated traffic”: sending automated queries to Google, for example to scrape results for rank checking, without permission. It violates Google’s terms of service, and it’s also why bulk rank checking from a single IP runs into captchas and blocks so quickly.
Fake clicks and engagement. Bots, click farms or paid “crowd” testers search for a keyword, click one result and stay on the page, hoping that CTR and engagement will lift it. Google has played clicks down for years, but testimony in the 2023 US antitrust trial described a system called Navboost that learns from how people click, so the idea refuses to die. I wouldn’t call it a reliable lever: it’s easy to fake badly, expensive to fake well, and it pollutes your own analytics.
Referral and “ghost” traffic spam. Fake visits or fake referrers that appear in your analytics to make you curious enough to open a link. It does nothing for rankings, it’s mostly noise — but it’s a good reminder not to trust raw traffic numbers.
9. Malware and malicious behavior
This one isn’t really SEO anymore, it’s cybercrime that uses search for distribution. Google’s policy covers malware, unwanted software (for example, apps that quietly change browser settings or leak private data), and any mismatch between what a user expects and what actually happens that puts their security or privacy at risk.
In practice, the malicious code usually ends up on a site without the owner knowing — through a hacked CMS, a compromised plugin or script, or an ad network. Google Safe Browsing can then flag the site, and visitors start seeing a red warning page in Chrome and other browsers. If you want the defensive side, read my article on hacked sites.
10. Deceptive functionality
Google calls it “misleading functionality”: a site that promises a tool, a service or a result it doesn’t really provide. Google’s own examples are a “free credit generator” for an app store that gives no credit, and simple tools (a PDF merger, a timer) that send visitors to deceptive ads instead of doing the job.
It works for SEO because tool pages rank for high-volume queries, and the visitor is worth money to whoever sits behind the ads. It stops working the way every tactic that depends on deceiving users stops working: people bounce, report the site, and Google’s systems — and sometimes people — notice.
11. Scaled content abuse
In March 2024 Google expanded its old “spammy automatically generated content” rule into scaled content abuse: producing many pages mainly to manipulate rankings, with little original value for users — no matter how they’re produced: AI, humans, or both.
Examples from Google’s documentation:
Mass-generated pages (with generative AI or similar tools) that add no value
Scraped or lightly transformed feeds and search results
Content stitched together from other pages without adding anything
Using several sites to hide the scale of the operation
The important detail: using AI is not the violation, scale without value is. I use AI in my own work every day, but a person decides what is worth publishing and checks every word that ships. That’s the difference between a tool and a content mill.
12. Scraped (stolen) content
Scraping means taking content from other sites, usually automatically, and republishing it to rank. Google’s examples:
Copying and republishing content with nothing added
Copying it and changing it slightly (swapping synonyms, automated rewrites)
Reproducing other sites’ content feeds without any unique benefit for the user
Sites that only compile or embed videos, images and other media from elsewhere
Besides the spam policy there is a legal side: copyright owners can send removal requests, and Google says a high volume of valid removal requests can demote other content on the same site.
13. Sneaky redirects
A sneaky redirect sends the visitor to a page different from the one the search engine saw, or to something that doesn’t match what they were looking for. It’s a close relative of cloaking. Typical examples:
The search engine sees one page, while real users are redirected somewhere else
Desktop users see a normal page, while mobile users are sent to a spam domain
Redirects themselves are fine and necessary: moving a site, merging pages, sending people after login. The violation is the deception.
Sneaky redirects are also a classic symptom of a hacked site, and they are sneaky for the owner too: the code often triggers only for visitors who arrive from Google, or only on mobile, so typing your own URL into the browser shows nothing wrong. If you suspect it, open your site from a search result on a phone.
14. Site reputation abuse (parasite SEO)
Site reputation abuse is when third-party content is published on an established, trusted site mainly to borrow that site’s ranking power. It’s the official name for what the industry has long called parasite SEO. Google’s examples: sponsored “payday loan” reviews on an educational site, or a low-quality casino advertising page on a medical site.
The policy was announced in March 2024 and Google began taking action on May 5, 2024. In November 2024 it clarified that the host site’s own involvement changes nothing: white-label deals, licensing and partial ownership don’t turn third-party content into first-party content.
What is not a violation: third-party content on its own, user-generated content like forums and comments, syndicated news and press releases, editorial columns, and affiliate content with properly marked links. It becomes a problem when the content is there to exploit the host site’s ranking signals.
The usual enforcement is a manual action that affects the abused section, not the whole site; the owner is notified in Search Console and can file a reconsideration request. In August 2026 Google changed how it applies this in the European Economic Area (from August 30): there the manual action no longer applies to search results, and the section may be separated in Google’s systems so it can rank on its own over time. Outside the EEA nothing changed. If you’re a publisher renting out sections of a strong site, this policy is aimed at you.
15. Thin affiliate pages
Affiliate marketing is not black hat. Thin affiliate sites are. Google defines them as affiliate pages where product descriptions and reviews are copied from the merchant, with no original content or added value. The typical picture is a cookie-cutter site copied across many domains or languages, differing only by the affiliate links.
What Google wants instead: price comparisons, original reviews, real testing, ratings, helpful navigation — anything a visitor can’t get by just opening the merchant’s page. And affiliate links should be marked with rel="sponsored" (see how to qualify outbound links).
16. UGC spam (user-generated spam)
This is spam that other people post on your site, in places meant for user contributions: comments, forum posts, profiles, free-hosting accounts, file uploads. The owner often doesn’t even know it’s there, but it’s on their domain, so it counts.
For spammers it’s cheap links and cheap pages on someone else’s authority. For site owners the defense is boring and it works: moderate (or require approval for) new content, add captchas or rate limits, close registration if you don’t need it, and mark user-submitted links with rel="ugc" (or nofollow) so they don’t pass ranking credit. Then clean up what’s already there.
Other methods usually called black hat
Google doesn’t give these their own line in the list above — most of them fall under link spam, the structured data guidelines or other policies — but in practice they come up all the time.
17. PBNs (satellites)
PBN stands for Private Blog Network: a group of sites, often built on expired domains (see section 3), whose main job is to link to the site that makes the money — the “money site” — with chosen anchors, while looking unrelated to each other. In some SEO circles they’re called “satellites”.
Google’s policies don’t mention PBNs by name. They don’t have to: a network built mainly to pass ranking credit is link spam by definition, and depending on how it’s made it can also be expired domain abuse or scaled content abuse.
The weak spot of every network is the footprints it leaves. When I look at a competitor, I’m not looking for magic — I’m looking for sites that share things they shouldn’t share: hosting and name servers, registrar and registration dates, the same theme and plugins, the same analytics or tag manager IDs, the same author names or contact details, the same linking patterns. Find one site and the rest often follow, which is how a whole network can lose its value at once.
18. Buying links
Link buying comes in several forms: ready-made placements on link marketplaces, paid guest posts, and “niche edits” — paying a site owner to add your link into an existing article that already has traffic and history.
Google’s rule is simple: links bought for ranking credit are link spam; links bought as advertising are fine if they carry rel="sponsored" or nofollow. And its systems look at it from both sides: they can flag the sites that buy links and the sites that exist mainly to pass links on.
19. Structured data (schema) abuse
Structured data tells Google what a page is about, and rich results (stars, prices, FAQs and so on) make a listing stand out. So people mark up things that aren’t there: fake reviews and ratings, content users can’t see, or a type that doesn’t match the page (instructions marked up as a recipe, for example).
Google’s structured data guidelines are clear: the markup must describe what users can actually see on the page. A violation can lead to a manual action, and the page loses eligibility for rich results. Google says that particular action doesn’t affect ranking itself, but the rich result — which was the whole point — is gone.
20. Auto-updating the “published” date
Fresh content ranks better for some queries, so some sites change the date on their pages automatically — every day, or on every visit — to look newly published or updated when nothing actually changed. You can see it in the visible date and in the datePublished / dateModified markup.
Google’s documentation on dates says they must describe when the page was published or updated, and that the visible date and the structured data should match. It’s a cheap trick, and the pattern — a date that changes while the content doesn’t — isn’t hard to spot. If you update an article, really update it, and let the date follow.
21. Site-level exploits and abuses
This is the broadest bucket: using a weakness in someone else’s website to get links or pages indexed. I won’t go into how it’s done, but the typical weak spots are worth knowing because they’re the ones you should check on your own site:
Outdated CMS, plugins and themes, and weak or reused passwords — the main way sites get hacked in the first place
Forgotten subdomains and DNS records that point to services nobody controls anymore
Internal search pages that anyone can fill with text and links and that search engines can index
Open redirects and other bugs that let outsiders send visitors through your domain
The common thread: the attacker borrows your site’s reputation, and you pay for it when Google notices.
Black, gray and white: where is the line?
“Black hat” isn’t a law, and it isn’t one thing. It’s whatever Google’s rules forbid — rules that Google changes, and that are enforced with different strictness on different sites. So a lot of real work happens in the gray zone. A few examples of how thin the line can be:
Links. A paid post with a sponsored label is advertising. The same post with a clean link and an optimized anchor is link spam.
Expired domains. A domain that continues what it was before, with real content for real visitors, is business as usual. A former school site filled with casino pages is expired domain abuse.
Content. AI-assisted articles that an expert checks and improves are fine. Ten thousand generated pages nobody reviewed are scaled content abuse.
Geo-targeting. Showing users in different countries different offers is common — it’s the example I used in the cloaking section. Showing Googlebot something different from what users see is cloaking.
What happens when a site gets caught
There are three typical outcomes:
Algorithmic demotion. Google’s systems (SpamBrain is the best known) stop counting the spam or lower the site. No message, just a drop in traffic that often coincides with a spam or core update.
Manual action. A person at Google reviews the site and applies a penalty. You see it in Search Console, under Security & Manual Actions, and after fixing the problem you can send a reconsideration request.
Removal from the index. Google says sites that violate its policies may rank lower or not appear in results at all.
One more thing worth knowing: Google says repeated attempts to get around its policies — for example, by starting over with new subdomains or new sites — can lead to restrictions on search features (like Top stories or Discover) or to removal of broader parts of a site. So “penalized? just rebuild a copy” is a risky plan.
How to protect your own site
You don’t have to do black hat to be hit by it. This is the short checklist I’d run on any site:
Keep the CMS, plugins and themes updated, use strong unique passwords and two-factor authentication.
Check Search Console regularly: the Manual actions and Security issues reports, and the page indexing report for pages you didn’t create.
Look at your site the way users and Google do: open it from a Google result on a phone, and use the URL Inspection tool to see what Googlebot gets.
Review your backlink profile from time to time. A sudden spike of low-quality links is worth a look (see the disavow article above).
Moderate user-generated content and label user links with rel="ugc".
Validate structured data and make sure it matches the visible page.
Check for forgotten subdomains and unused DNS records.
FAQ
Is black hat SEO illegal?
Breaking Google’s rules isn’t a crime — the penalty is lower rankings or removal from the index. But some of the methods are illegal in themselves: hacking a website, distributing malware, copying copyrighted content, deceiving users to get their money or data. Those risks have nothing to do with Google.
Does black hat SEO still work in 2026?
Sometimes, for a while. Spam policies exist to shorten that “while”, and Google keeps getting better at spotting patterns. A business built on tactics that can be banned overnight has a very short planning horizon.
Can a site recover after a penalty?
Often, yes. For a manual action you fix the problem, explain what you did in a reconsideration request, and wait for the review. For an algorithmic demotion there’s no request: you fix the cause and wait for the systems to re-evaluate the site, which can take a while. Cleaning up a link profile or a hacked site is a project of its own; I cover both in separate articles.
Is using AI to write content black hat?
No. Google’s rule is about purpose and scale, not tools: useful AI-assisted content that a person reviewed is fine; mass-produced pages made to manipulate rankings are scaled content abuse.
So, is SEO really broken?
I started this article by saying that SEO is completely broken, and this is what I meant. The rules are written by one company, they change, and they’re enforced mostly by algorithms that see patterns, not intentions. That leaves a gap between what the guidelines say and what actually ranks — and black hat lives in that gap. Examples of most of these tactics working somewhere, for some time, aren’t hard to find.
That doesn’t make it a good strategy, and I’m not telling you to use any of it. I write about it because to protect a site, to clean up a link profile, or just to understand why a competitor is outranking you, you need to know how these things are done and where Google draws the line.