Under Attack? My Negative SEO Investigation

Case study · Spam backlinks · October 2026

1,100 Spam Backlinks in 8 Days: Should You Disavow?

Short answer: probably not. My SEO consultancy site picked up more than 1,100 spammy backlinks in a week, and its impressions dropped at the same time. Here's the data that showed the links weren't the cause, and the checks to run before you touch the disavow tool.

Thousands of identical web pages linking to a single website

Quick answer

Should you disavow spam backlinks? Usually not. A sudden flood of junk links is most often an automated link farm, and Google says it ignores most of those links on its own. Disavow only if you built or bought bad links, got a manual action, or can tie a drop in clicks to the links after ruling out everything else. In this case, the "attack" lined up with a Google spam update and a breakdown in SEO tracking data, and those explained the drop better than the links did.

1,188referring domains on Oct 13, up from about 104 in August
97.8%of them scored 30+ on DataForSEO's spam score
1link per domain, all from one copied template
0hidden links, cloaking or injected pages on my sites

What happened: 1,100 spam backlinks in eight days

Between October 5 and October 13, 2026, optimizationtheory.com went from 507 to 1,188 referring domains. Almost every new one was a throwaway site on a cheap TLD, sending a single link.

The flood had been building since August, when the site had about 104 referring domains. By the end of September it had 298. Then it took off: backlinks went from 1,090 to 1,841 in those eight days, and 943 referring domains were first seen in October. New ones were still showing up on the day I wrote this.

At the same time, Search Console impressions fell 45% over 28 days, from 6,051 to 3,305, dropping to about 50 a day on October 9 and 10. It looked like textbook negative SEO. Two of my other sites were hit too, but differently:

Same weeks, three different storiesHeadline metric per site, the 28 days to Oct 10 against the period beforeoptimizationtheory.comGSC impressions, 28 daysPrior 28 days: 6,0516,051Prior28 daysSep 13 to Oct 10: 3,3053,305Sep 13 toOct 10down 45%Likely partly measurement (Finding 2)vuss.ioGSC average position (higher = better)#1#5#10#15Before: 11.5Recent: 5.311.55.3BeforeRecentimpressions ~2xGains began Sep 30infographi.acDataForSEO ranked keywordsBefore : 4141BeforeOct 13 : 2121Oct 13down 49%Tool data, unreliable since Sep 13
Sources: owner's Search Console and DataForSEO data, pulled October 5 and 13, 2026.
  • vuss.io got a smaller wave of about 40 casino and gambling domains, yet its average position improved from 11.5 to 5.3 and impressions roughly doubled.
  • infographi.ac got almost no spam links, but its ranked keywords in DataForSEO halved from 41 to 21.

The domains themselves were the giveaway. They sat on six cheap TLDs, and their names read like SEO tools: onlinebacklinkgenerator.shop, aibacklinktool.space, bestlinkbuildingstrategies.site, backlinksearch.website, serpcheckeronline.site, dacheckers.online.

Six cheap TLDs supply 85% of the referring domainsoptimizationtheory.com referring domains by TLD, DataForSEO, Oct 13, 2026 (total 1,188).website.website: 183 domains183.store.store: 176 domains176.shop.shop: 175 domains175.site.site: 168 domains168.space.space: 163 domains163.online.online: 149 domains149Other TLDsOther TLDs: 174 domains174The six named TLDs account for 1,014 of 1,188 domains. "Other" is the remainder.
Source: owner's DataForSEO backlink export, October 13, 2026.
97.8% of referring domains score 30 or higher on spamDataForSEO spam score, optimizationtheory.com referring domains, Oct 13, 2026Spam score 30+: 1,165 domainsUnder 30: 26 domains1,165 domains score 30+26 under 30The domain's overall backlink spam score rose from 27 to 38 between the Oct 5 and Oct 13 pulls.
Source: owner's DataForSEO export. Spam score is DataForSEO's metric, not Google's.

On October 13 I submitted disavow files covering 1,164 domains for optimizationtheory.com and 44 for vuss.io. As you'll see, that probably wasn't necessary.

Negative SEO attack or link farm? How to tell

Every linking domain I opened was a copy of the same page. That's the signature of an automated link farm hitting thousands of sites at once, not a competitor targeting you.

An assembly line stamping out identical spam web pages

I opened four of the spam domains on October 13 and compared them side by side:

Footprint strongbacklinks .online backlinksearch .website dacheckers .online onlinebacklink generator.shop
"Money Robot" header linking to itself ✓ ✓ ✓ ✓
"Grow Organic Traffic For FREE" hero ✓ ✓ ✓ ✓
Long generic SEO article on the domain's keyword ✓ ✓ ✓ ✓
/docs/ sub-articles and a /dir/1 "index" page ✓ ✓ ✓ ✓
Pricing, Testimonials, Contact are dead "#" links ✓ ✓ ✓ ✓
Owner, address, copyright or dates None None None None

What Money Robot backlinks look like

Money Robot Submitter is long-running commercial software that automates link building across web 2.0 blogs, profiles, wikis and bookmarking sites, and reviewers have long called it spam (Tech Business News, Jun 2026). The domains linking to me carried its name, but I couldn't confirm whether Money Robot runs them or a clone borrows the branding. Either way, the pattern is the same: keyword-named domains, a copied landing page, and links to target sites tucked into /dir/ directory pages.

Other site owners have reported the same footprint. In July 2026 a BlackHatWorld member described brand-new sites getting "blasted" by SEO-keyword .shop domains like seoboost.shop and ranktools.shop, with operators who "spin up 50 new domains" faster than anyone can disavow (BlackHatWorld). The agency Hallam documented 28 similar .shop domains across several industries (Hallam).

Signs of a link farm

  • One link per domain
  • Cheap TLDs: .shop, .site, .online, .store, .space, .website
  • The same page template on every domain
  • No owner, address or contact details
  • The same domains reported by other site owners

Signs of a targeted attack

  • Exact-match or adult anchor text aimed at your money pages
  • Links timed with a real drop in clicks
  • Fake removal requests or scraped copies of your content
  • Extortion emails offering to "stop" the links
  • Links only to you, not a whole niche

My readBecause the domain names are SEO keywords, the farm probably seeds its directories with SEO-related sites so the pages look on-topic. That would explain why my SEO consultancy got about 1,100 domains while my infographic site got almost none.

Did the spam backlinks hurt rankings? What the data showed

Probably not. The impression drop started on September 13, weeks before most of the links arrived. That's the same day Google started blocking SEO tools far more aggressively.

A broken analytics dashboard behind a shield blocking a scraper robot

Around September 13, Google began blocking rank trackers and scrapers much more effectively. Nozzle said about 80% of its scraping attempts were blocked, and that DataForSEO was hit the same way (SER, Sep 18, 2026). Nozzle also found that from September 20, some results pages served to trackers were falsified: they looked normal but were missing most of the usual sites.

Poisoned results pages hide tracked sites about 20x more oftenNozzle rank-tracking research, data through Oct 9, 2026Usual page-1 sitesstill presentNormal: 88.5%88.5%Poisoned: 20.7%20.7%Tracked site pushedout of the top 10Normal: 2.7%2.7%Poisoned: 58.6%58.6%Normal results pagesPoisoned results pages (8.0% of all, from Sep 13)
Source: Nozzle, Poisoned Google SERPs, October 10, 2026.

That matters in three ways:

  • Tool data broke. infographi.ac's drop from 41 to 21 ranked keywords lines up with DataForSEO losing most of its Google data, not with a real ranking loss. Keyword counts can fall sharply while real traffic holds (Search Engine Land, Sep 24, 2026).
  • Search Console lagged. The performance report ran more than 64 hours behind on October 11 (SER, Oct 11, 2026), so my 50-a-day impressions on October 9 and 10 were likely incomplete.
  • Bot impressions may have vanished. When Google blocked scrapers in September 2025, about 88% of 320 sites lost Search Console impressions (SER, Sep 22, 2025). An SEO site ranks for exactly the terms rank trackers scrape. HypothesisThe test: if clicks held and the lost impressions sit at deep desktop positions, it's a measurement artifact.

Here's how the link flood lines up against everything else that happened:

Junk links quadrupled after the tools broke and the spam update began optimizationtheory.com referring domains against outside events, Aug to Oct 2026 Aug 1Aug 15Sep 1Sep 15Oct 1Oct 15 Google's scraper blocking intensifies Referring domains DataForSEO pulls Aug (approx.): 104 referring domainsSep 30: 298 referring domainsOct 5: 507 referring domainsOct 13: 1,188 referring domains 1042985071,188 Outside events and data windows Aug spam updateScraper blockingGSC study windowPoisoned SERPsSep spam updatevuss.io gainsGSC 64h delayDisavow filed Aug 18 to Aug 21Sep 13 to Oct 13Sep 13 to Oct 10Sep 20 to Oct 13Sep 24 to Oct 8Sep 30Oct 11Oct 13
Owner's DataForSEO pulls (August value approximate) and the dated reports listed in Sources. Aug 1 to Oct 15, 2026.

The jump from 298 to 1,188 referring domains came after the scraper blocking (Sep 13) and the September spam update (Sep 24) had already started. Blaming the links means ignoring both.

Other false alarms that look like an attack

Two more things hit at the same time and made the situation look worse than it was: fake analytics traffic and a Google spam update.

GA4 ghost traffic from China, Singapore and the Netherlands

Ghost visitors entering analytics without passing through the website

A .txt file on vuss.io logged 71 pageviews, and a .txt file can't run the Google Analytics tag. That points to fabricated hits sent straight to GA4's collection endpoint with the site's public Measurement ID (Simplytics, Sep 2026). Google Product Experts have called the China and Singapore traffic "inauthentic, non-human traffic" (Stan Ventures, Dec 2025).

Three countries sent about a quarter of vuss.io sessions as engagement halvedGA4, vuss.io, same window as the Search Console comparisonSessions by countryabout 26% of all sessions combinedNetherlandsNetherlands: 286 sessions286SingaporeSingapore: 134 sessions134ChinaChina: 65 sessions65Direct-traffic engagement ratefell by halfBefore: 41%41%BeforeAfter: 20%20%AfterA .txt file that cannot run the GA tag also logged 71 pageviews.
Source: owner's GA4 property for vuss.io.

Fix: turn on GA4's hostname "Include" filter, added September 21, 2026 (Search Engine Watch), build exclusion segments, and compare against Cloudflare logs. Visits Cloudflare never saw are ghost hits, and no firewall rule can stop them.

The Google spam updates

Storm fronts sweeping across a landscape of search results

Google ran two spam updates during the window: August 18 to 21, and September 24 to October 8. Google said neither targeted link spam. The September one ran about 14 days, far longer than usual (SER), with impact waves on September 25 to 27, September 30 and October 4 to 6. vuss.io's gains started September 30, right on the second wave.

The September spam update ran five times longer than August'sRollout length of each 2026 Google spam update, in daysMarch: ~1 day (Mar 24 to 25)~1 dayMarchMar 24 to 25June: ~2 days (Jun 24 to 26)~2 daysJuneJun 24 to 26August: 2.7 days (Aug 18 to 21)2.7 daysAugustAug 18 to 21September: 13.7 days (Sep 24 to Oct 8)13.7 daysSeptemberSep 24 to Oct 8Google said the August and September updates did not target link spam. March and June lengths are approximate.
Sources: Relevant Audience, SER (August), SER (September).

How to check your site for spam backlinks and hacks

Before blaming outside links, rule out problems on your own site. I scanned all three sites from the outside and found no hack, but one open door.

A website being scanned clean, with one small open door highlighted
Check How optimization theory.com vuss.io infographi.ac
Cloaking Compare a page fetched as a browser, as Googlebot and with a Google referrer Clean Clean Clean
Hidden outbound links Search page HTML for links inside display:none or off-screen blocks None None None
Injected scripts Look for eval, atob, fromCharCode or stray iframes None None None
Spam pages in sitemaps Scan sitemap URLs for foreign hosts or casino slugs 0 of 126 0 of 118 0 of 256
Doorway pages Request a made-up spam URL; it should 404 404 404 404
Search page spam Load /?s=online+casino+slots and check for noindex noindex noindex Indexable
Exposed users Open /wp-json/wp/v2/users 1 user 1 user Incl. "admin"

Why the search page matters: spammers link to your site's search URLs with casino terms in the query, hoping Google indexes pages of spam text on your trusted domain. If your search results aren't set to noindex, fix that first. It's a bigger risk than any number of junk referring domains.

An outside scan can't see server files, the database or cloaking aimed only at Google's real crawler IPs. Cover that with URL Inspection's "View crawled page" in Search Console, plus the Security Issues and Manual Actions reports.

Should you disavow spam backlinks? What Google says

Google's guidance has been consistent: most sites never need the disavow tool, and spam links you didn't build are mostly ignored.

Disavowing so-called toxic links is a "billable waste of time."

John Mueller, Google, Bluesky, Dec 17, 2024 (SER)

"There's a lot of hate on the disavow tool here - but it has its place & time, it does have effects." And: "disavowing a handful of links is not going to make your site drop out of search."

John Mueller, Reddit, reported Sep 2, 2026 (SER)

Spam links "would not have any effect, positive nor negative."

John Mueller on comment spam links, Bluesky, Jan 18, 2026 (PPC Land)

Disavow when

  • You bought links or ran link schemes yourself
  • Search Console shows a manual action for unnatural links
  • Clicks (not just impressions or tool metrics) dropped and nothing else explains it

Don't bother when

  • The links are from an obvious farm you had nothing to do with
  • Only third-party spam scores or "toxic" ratings went up
  • Rankings and clicks are holding steady

Lesson from my case: I disavowed 1,164 domains before checking clicks. That was a reflex, not a diagnosis. The one real risk of disavowing is catching legitimate links in the net, which is the situation where Mueller says the tool "does have effects."

How to disavow backlinks (if you really must)

A stethoscope on a laptop of graphs with scissors set aside
  1. Export your backlinks from Search Console's Links report and your link tool of choice.
  2. Filter to the farm's footprint: cheap TLDs, one link per domain, first seen in the spike window.
  3. Check the list by hand for any real sites, partners or directories you want to keep.
  4. Write a plain .txt disavow file with one entry per line, using domain-level entries.
  5. Upload it in Google's Disavow Links tool for the correct property. A new upload replaces the old file, so keep a master copy and add to it.

Disavow file example

# Spam link farm, Money Robot template, Oct 2026
# Reviewed by hand on 2026-10-13
domain:onlinebacklinkgenerator.shop
domain:aibacklinktool.space
domain:bestlinkbuildingstrategies.site
domain:backlinksearch.website
domain:serpcheckeronline.site
domain:dacheckers.online

Mueller has said you can disavow a whole TLD with an entry like domain:.shop "if you're sure that it's what you want to do" (ALM Corp, Mar 2026). Be careful: that removes every link from that TLD, including real ones.

Checklist: what to do when spam backlinks spike

Diagnose before you disavow. Most of the work is confirming what your data actually means.

  • Check clicks in Search Console, not just impressions, and compare desktop with mobile.
  • Line up the drop's start date with Google updates and known data outages before blaming links.
  • Open five or ten of the spam domains. If they share one template, it's a farm, not a targeted attack.
  • Don't trust third-party rank or keyword counts from mid-September 2026 onward until tracking stabilizes.
  • Make sure your site search results are set to noindex.
  • Run URL Inspection on your homepage and key pages, and check Security Issues and Manual Actions.
  • Filter GA4 ghost traffic with the hostname Include filter and exclusion segments.
  • Only then decide whether a disavow file is worth it, and review it by hand.

Stop worrying about

  • The raw count of junk referring domains
  • Third-party spam scores, which are the tool's opinion, not Google's
  • Updating your disavow file daily as new domains appear

Do worry about

  • A real drop in clicks or conversions
  • Indexable search pages or unknown admin users
  • Manual actions or security warnings in Search Console

Spam backlinks FAQ

Do toxic backlinks hurt SEO?

Rarely. Google's systems are built to devalue spam links rather than penalize the site they point to. In this case, about 1,100 spam domains arrived in eight days, and the evidence points to measurement problems, not the links, behind the drop in impressions.

How do I get rid of spam backlinks?

You usually can't remove them, because the sites hosting them have no contact details. Confirm the footprint, check whether clicks actually fell, and if you must act, add domain-level entries to a disavow file in Search Console.

Is negative SEO real?

Attempts happen, but most sudden spam link floods are automated link farms hitting many sites at once. One link per domain, cheap TLDs and an identical template point to a farm, not a targeted attack.

What is a good spam score for backlinks?

Spam score is a third-party metric from tools like Moz or DataForSEO, and Google doesn't use it. Here, 97.8% of the new domains scored 30 or higher. That confirmed they were junk, but said nothing about whether Google counted them.

What does disavow mean in SEO?

Disavowing tells Google not to count specific links or domains when assessing your site. You do it by uploading a .txt file in Search Console's Disavow Links tool. Google treats it as a strong suggestion, and most sites never need it.

How do I check for spam backlinks?

Search Console's Links report shows the domains Google sees. Tools like DataForSEO, Ahrefs or Semrush add first-seen dates and spam scores. Sort by first-seen date: a farm shows up as hundreds of cheap-TLD domains appearing in a few days.

Sources and limitations

This is one owner's network over about ten weeks. I opened four of about 1,100 spam domains by hand; the /dir/ pages didn't load, so where the links sit is inferred. The bot-impression explanation is a hypothesis built on the 2025 precedent. Site data came from my own DataForSEO, Search Console and GA4 accounts, pulled October 5 and 13, 2026. All sources below were opened October 13, 2026.

Date Source Used for
Oct 12, 2026 SER: Third-Party Google Rank Tracking Struggling Now Tool data notices
Oct 11, 2026 SER: Search Console Performance Report Extremely Delayed 64-hour GSC delay
Oct 10, 2026 Nozzle: Poisoned Google SERPs Falsified results served to trackers
Oct 8, 2026 SER: September 2026 Spam Update Done Update dates and waves
Sep 30, 2026 Simplytics: GA4 ghost traffic How ghost hits work
Sep 29, 2026 Search Engine Watch: GA4 hostname filter Include filter
Sep 24, 2026 Search Engine Land: Google's /goto change Keyword-count distortion
Sep 18, 2026 SER: Google More Successful In Blocking Scrapers Sep 13 blocking
Sep 2, 2026 SER: The Disavow Tool Does Have Effects Mueller on disavow
Aug 21, 2026 SER: August 2026 Spam Update Done Update dates
Jul 11, 2026 BlackHatWorld: automated spam backlinks on Day 1 Same footprint, earlier
Jun 2, 2026 Tech Business News: Money Robot review How Money Robot works
Mar 10, 2026 ALM Corp: TLD disavow Mueller on domain: entries
Jan 19, 2026 PPC Land: comment spam Mueller on spam links
Dec 16, 2025 Stan Ventures: GA4 bot traffic Google Product Experts statement
Sep 22, 2025 SER: Impact of blocking bots on GSC 2025 precedent
Dec 17, 2024 SER: Disavowing toxic links "Billable waste of time"
Undated Hallam: spam referring domains .shop SEO-keyword domains
Undated Relevant Audience: 2026 update list Earlier 2026 updates

Spam domains named in this article are listed, not linked. © 2026 Optimization Theory. Seeing a spam link flood on your own site? Get in touch.