Skip to content
There are two unreliable narrators in SEO. The first is Google, whose public statements are carefully worded, frequently incomplete, and shaped by the fact that it would prefer you didn’t manipulate its rankings at all. The second is the SEO community, whose collective wisdom is a mixture of hard-won experience, outdated folklore, superstition, and confident nonsense repeated until it sounds true. Most PBN advice you’ll encounter comes from one of these two sources, which means most PBN advice is, to some degree, wrong.
So rather than pick a side, this is an audit. For each of the most common claims in PBN and link building, we’ll lay out what Google says, what the community says, and then — the only column that matters — what’s actually true once you account for the fact that both of the other two are talking their own book. The honest answer is usually that everyone is a little bit wrong, including Google. Here’s where the truth actually sits.
Claim 1: ‘Links Don’t Matter Much Anymore’
What Google says: Links have been steadily downgraded in Google’s public messaging. Gary Illyes said at Pubcon that links aren’t a top-three ranking factor and haven’t been for some time. In March 2024 Google quietly removed the word ‘important’ from how its spam documentation described links. Illyes later said Google needs ‘very few links’ to rank a page, and John Mueller has said he expects the weight on links to keep falling over time.
What the community says: Pure disinformation. Google knows links are the most abusable signal it has, so of course it tells everyone they don’t matter — to discourage manipulation while continuing to rely on them heavily. Operators point at the SERPs and see thin-content pages with strong link profiles outranking better content every day, and conclude links are as decisive as they ever were.
What’s actually true: Both are spinning. Google is technically not lying — links genuinely are a smaller share of the ranking picture than in 2010, because Google now has hundreds of signals and far better content understanding. But the framing is self-serving: ‘not top three’ and ‘we need very few links’ quietly imply ‘don’t bother,’ which is not the same statement and not true. The community is right that links still move rankings hard in competitive niches, but wrong to treat the downgrade as pure theatre — the weighting genuinely has shifted, and link-only strategies that ignored content worked far better five years ago than they do now.
The accurate position: links remain one of the most powerful signals you can directly influence, but they’re now necessary-not-sufficient — a strong profile pointing at weak, derivative content underperforms in a way it didn’t used to. We went deeper on this exact gap in separating Google’s claims about links from what operators actually observe, because it’s the single most consequential thing to get right.
Claim 2: ‘PBNs Are Dead’
What Google says: Google has said for over a decade that link schemes including private blog networks violate its guidelines, that it began actively de-indexing PBNs as far back as 2014, and that it tries to algorithmically ignore links it identifies as manipulative. The official line is that PBNs don’t work and will get your sites penalised.
What the community says: Split down the middle. One camp declares PBNs dead every year — the detection is too good, the cost of doing it properly is too high, you’re better off with white-hat link building. The other camp quietly keeps running networks that quietly keep working, and views the ‘PBNs are dead’ chorus as either cope from people who did it badly, or deliberate misdirection to thin out the competition.
What’s actually true: The phrase ‘PBNs are dead’ is lazy. What’s actually dead is the 2014-era cheap PBN: identical themes, shared hosting, thin spun content, exact-match anchors, a single outbound link per site. That model is trivially detectable now and genuinely doesn’t work. What isn’t dead is a network of sites that are individually believable — properly hosted, genuinely written, topically coherent, behaving like real websites. The detection got dramatically better, which killed the cheap version and raised the cost and skill floor for the version that still works.
So the truth is neither ‘dead’ nor ‘fine.’ It’s that the bar moved. The networks performing in 2026 are the ones built to look like building networks that behave like real sites rather than link farms rather than like link farms — and the gap between those two things is exactly the gap between the dead version and the living one.
Claim 3: ‘Exact-Match Anchor Text Will Get You Penalised’
What Google says: Anchor text stuffing — overloading your backlinks with exact-match commercial keywords — is treated as a manipulation signal. In most cases Google says it simply ignores these links, but in extreme cases a high volume of them can contribute to a manual action.
What the community says: Two contradictory camps, loudly. One treats exact-match anchors as radioactive — keep them under some magic percentage or face instant doom, run every profile through a ‘toxic anchor’ checker, panic at the first money-keyword anchor. The other camp says anchors are the whole point of a PBN — the entire reason you control the link is to control the anchor, so use the keyword you want to rank for.
What’s actually true: Both camps are reacting to a real thing and overshooting in opposite directions. Exact-match anchors are genuinely a strong manipulation signal — an unnatural concentration of them is one of the clearest patterns Google’s systems and manual reviewers look for. But the panic-driven ‘never use them’ position throws away the legitimate ranking value that relevant anchors carry, and the ‘anchors are everything’ position walks straight into the signal Google is best at catching. The truth is distributional: a natural profile contains a mix dominated by branded, URL, and natural-phrase anchors, with exact-match as a small minority. The danger was never the existence of a money anchor; it’s the ratio.
This is why anchor strategy is about the shape of the whole profile, not a rule about individual links — the subject of building a natural anchor text profile rather than chasing exact match. Chasing a percentage misses the point; building a profile that looks like one a real site would naturally attract is the point.
Claim 4: ‘You Should Disavow Toxic Links’
What Google says: Google provides the disavow tool but cautions repeatedly against overusing it, says most sites never need it, and states that it already ignores the vast majority of spammy links automatically. Mueller has said on the record, more or less forever, that the average site should not be disavowing links.
What the community says: A whole industry of ‘toxic link’ checkers and monthly disavow ‘maintenance’ says otherwise — that you must routinely audit your profile, identify toxic backlinks by some tool-generated score, and disavow them to stay safe. It’s sold as basic hygiene.
What’s actually true: Here the community is mostly wrong and Google is mostly right, which doesn’t happen as often as Google would like you to think. Controlled tests have repeatedly shown that disavowing tool-flagged ‘toxic’ links does close to nothing positive — and can actively hurt, because you may disavow links that were quietly helping you. Google really does ignore most low-quality links algorithmically, so disavowing them just formalises something already happening. The genuine exceptions are narrow: a confirmed manual action citing unnatural links, or a real negative-SEO attack. Outside those, routine disavow ‘maintenance’ is a service being sold to you for a problem you almost certainly don’t have.
The one context where disavow genuinely matters — a manual action — is also the one most people never face and most misunderstand. What that actually looks like, and how recovery really works, is covered in what a real manual action looks like and how recovery works.
Claim 5: ‘Google Can Instantly Detect Any PBN’
What Google says: Google projects total confidence: it has SpamBrain, it finds footprints, it neutralises link schemes at scale, resistance is futile. The messaging is deliberately designed to make manipulation feel hopeless.
What the community says: Also split. The fearful camp treats Google as omniscient — assume every footprint is already detected, every network already flagged. The dismissive camp points out that obvious spam networks rank for years untouched and concludes detection is mostly bluff.
What’s actually true: Google’s detection is genuinely sophisticated and genuinely better than most operators assume — but it is pattern-based, not omniscient, and it is not instant. SpamBrain is very good at finding footprints: shared hosting, identical configurations, unnatural link patterns, the signatures of networks built cheaply and at scale. It is far weaker against sites that don’t exhibit those patterns, because there’s no footprint to find. The ‘omniscient’ framing is partly Google’s own PR; the ‘it’s all bluff’ read is wishful thinking that ignores how many networks quietly get neutralised. The truth is that detection is a question of footprints and patterns — minimise them and you’re genuinely hard to catch; leave them and you’re genuinely easy.
Understanding what the detection systems actually key on — rather than treating them as either magic or bluff — is the whole game, and it’s the subject of how SpamBrain actually detects manipulative link patterns.
The Pattern Behind All Five
Step back and the same shape appears in every audit above. Google’s statements are technically defensible but framed to discourage the behaviour it’s describing — true-ish, in the direction that serves Google. The community’s wisdom contains a real observation distorted by either fear or wishful thinking, usually overshooting into a rule that’s easier to follow than to justify. And the actual truth almost always sits in a more boring, more conditional middle: it depends on ratios, on execution quality, on footprints, on whether the underlying content is real.
That middle is less shareable than ‘links are dead’ or ‘just use exact-match anchors,’ which is exactly why the extremes spread and the truth doesn’t. Slogans travel; nuance doesn’t. But the operators who perform consistently are the ones who’ve stopped taking either narrator at face value — who treat Google’s guidance as evidence-to-be-weighed rather than gospel, and treat community consensus as folklore-to-be-tested rather than fact.
The throughline of every honest answer in this audit is the same: the cheap, patterned, derivative version of each tactic is what’s dead or dangerous, and the genuine, well-executed, real-looking version is what still works. That’s not a coincidence — it’s the whole principle behind network content that genuinely earns its place and behind building networks that earn their place rather than gaming a system that’s better at catching gamers than it used to be. The full framework is laid out in the complete PBN guide for 2026, and if you want a straight answer about a specific piece of advice you’ve been given — from Google or from a forum — free personalised SEO and PBN advice is available, and we’ll tell you which of the three columns it really belongs in.