John Mueller was asked point-blank on Bluesky if a penalized site's outbound links pass "poor signals." His answer settles an old SEO myth.
"Bad neighborhood" is one of the oldest pieces of SEO folklore still in circulation: the idea that a link from a spammy or penalized site does not just fail to help you, it actively drags you down, the same way living next door to a meth lab lowers your property value. It is old enough that Google's disavow tool was partly built to give people a way to distance themselves from exactly this fear. The disavow tool is functionally dead now, but the fear it was built to soothe never fully went away, and it just got a direct, current answer.
The question someone finally asked outright
In April 2026, an SEO going by @searchassistance.co.uk asked John Mueller on Bluesky the exact mechanism people usually only imply: "If a site has a link penalty are the outbound links from that site devalued? Or do they have the ability to pass on poor signals, ie bad neighbours?" It's a fair question with two very different possible answers. One is that a flagged site's outbound links simply stop counting, worth zero. The other is that they actively transmit something negative to whatever they point at, worth less than zero.
Mueller's reply, as reported by both Search Engine Journal and Search Engine Roundtable: "I'm not sure what you mean with 'has a link penalty', but in general, if our systems recognize that a site links out in a way that's not very helpful or aligned with our policies, we may end up ignoring all links out from that site. For some sites, it's just not worth looking for the value in links."
Ignored is not the same as harmful
Read that answer twice, because the two outcomes it describes are easy to conflate and they are not the same thing. A link that's ignored contributes nothing, neither credit nor damage. A link that actively passes a poor signal would make you worse off for having received it, the digital equivalent of guilt by association. Mueller describes only the first mechanism. Google's systems can choose not to bother evaluating links from a site once they've decided that site's outbound linking isn't helpful or policy-aligned. What he does not describe, and what nothing in his answer supports, is Google reaching downstream and marking the recipient site as tainted because of who linked to it.
That distinction is the whole ballgame for anyone who has ever worried about "bad neighborhoods." A link from a spam-riddled directory or a penalized site is not evidence against you. At worst, it's a link Google already decided to stop counting, which puts you back exactly where you'd be if that link never existed. It does not put you in a hole.
Where the fear comes from
The bad-neighborhood idea isn't pure invention. It traces back to real link-graph research: algorithms that use a seed set of known-trustworthy sites and measure how many hops away a given site sits from that set, on the logic that good sites cluster with good sites and spam clusters with spam. That's a real pattern Google's systems can use to help identify what's spammy in the first place. But using link relationships to spot spam is a different mechanism from using them to spread a penalty. One is classification. The other would be contagion. Mueller's answer confirms Google does the former and not, as far as the outbound-link question goes, the latter.
The confusion has a mundane explanation, too: a specific idea gets written up once, other writers paraphrase it without the original nuance, and a few rounds of that compression later, "sites in a bad neighborhood get flagged by association" is treated as settled fact by people who never read the source material it was extrapolated from.
What this changes about vetting a link
None of this is an argument for accepting links from anywhere. A backlink placement is still worth vetting on its own merits, whether it sends real referral traffic, whether it's topically relevant, whether the page looks like it exists for readers rather than for other advertisers. But the vetting question is "is this link doing anything useful for me," not "could this link actively hurt me by association." Those are different risk models, and treating the second one as real leads to wasted time auditing link neighborhoods for a contamination effect that Mueller's answer says doesn't work that way.
It also reframes what a link penalty from Google costs the site that has one. It's not a weapon that keeps working against everyone that site links to after the fact. It mostly just means that site's outbound links, all of them, stop being worth chasing, which is a cost the flagged site pays, not the sites it links to.
Where this differs from what's already on this blog
The Google-said-low-quality-backlinks-are-useless post covers a different, broader question: whether Google ever said incoming link quality doesn't matter at all. This is narrower and mechanistic: whether a link from a specific kind of flagged site can actively transmit harm to whatever it points at, versus simply not counting. Those are different claims with different answers, and conflating them is part of how the bad-neighborhood myth keeps outliving the disavow-tool era it came from.
A directory listing doesn't carry any version of this risk either way. Inclusion in a well-kept directory is either a real, evaluated placement or it isn't; there's no mechanism by which sitting near another listing could reach across the page and poison a ranking.
