What an expired domain actually is
Strip the marketing and an expired domain is one thing: a registration that lapsed because the previous owner stopped paying, and that has now cycled back into the available pool, often through a drop-catch auction. Everything that makes it interesting sits outside the name itself. It is the referring-domain profile, the anchor distribution, the indexation history, and the topical footprint that accumulated over the years it was live. You are not buying a word with a TLD, you are buying a graph position.
That distinction matters because the market sells the wrong thing. Most listings lead with a Domain Rating or Trust Flow number, which is the single most gameable signal in the entire transaction. A DR 60 built on three hundred links from one PBN is worth less than a DR 28 built on forty editorially earned referring domains. We treat the headline metric as a filter to reject, never as a reason to buy. The real read is qualitative: who linked, why, and whether those links would still be placed today.
The other half of the asset is liability. Every expired domain carries the history of whatever its last owners did with it. A domain that spent two years as a redirect farm or a thin-affiliate site is not neutral, it is pre-poisoned. The reason expired domains command a premium over fresh registrations is the hope of inherited trust, and the reason most of them disappoint is that the inheritance is negative.
How trust transfer works in 2026
The mechanism people imagine, that authority flows automatically from an old domain to whatever you put on it, was never quite true and is decisively less true now. In March 2024 Google added « expired domain abuse » to its spam policies as a named category, describing the practice of buying an expired domain and repurposing it to host low-value content with the primary goal of manipulating rankings via the prior reputation. That was not a vague algorithm tweak, it was a written policy line that gives Google explicit cover to act, manually or algorithmically.
So in 2026 the operative model is conditional, not automatic. Residual trust survives when there is continuity: the new site covers a topic adjacent to the old one, the content is genuinely useful, and the link graph still makes sense in context. It evaporates, or inverts into a penalty, when the domain is hollowed out and pointed at an unrelated commercial vertical. Topical continuity is the variable that decides which path you are on, and it is the one most operators ignore because it is harder to fake than a metric.
Measurement in practice means three layers. First, the Wayback Machine to reconstruct what the domain actually published and whether it was a real site or a spun shell. Second, a live backlink crawl in Ahrefs or Semrush to see which referring domains still resolve and still link, because a profile that looked strong in a cached export can be half dead today. Third, an index check: a quick site query and a manual look at whether Google still surfaces anything, which tells you whether the domain is in good standing or already sandboxed. If you want a sense of how this overlaps with classic placement tactics, our entry on the practice of inserting links into existing aged content covers the same trust-continuity logic from the link side.
Where it fits in a netlinking operation
There are two legitimate uses and one that we steer clients away from. The first legitimate use is the rebuild: you acquire an expired domain whose backbone matches a niche you want to operate in, you restore or recreate genuinely useful content, and the site becomes a standalone publisher that can link out naturally over time. This is slow, it costs editorial effort, and it is the only model that ages well. It is also, not coincidentally, close to how an owned editorial network is supposed to work rather than a rented one.
The second use is the satellite or feeder: a smaller rebuilt site in a related language or market that sends a handful of contextual links toward money pages. This is defensible when the feeder is a real site with its own traffic and its own reason to exist. It becomes a footprint liability the moment it is obviously a thin shell built only to link, which is exactly the pattern the 2024 policy targets.
The use we push back on is the bare 301 redirect of an expired domain straight into a client's money site. It is the fastest way to move equity and the fastest way to import a penalty, because you inherit the entire history with zero editorial buffer. If the domain was clean it can work, but you rarely know it was clean until after the redirect has already passed whatever it was carrying. When a client wants authority moved into their site, a calibrated campaign of contextual placements is lower variance than a redirect gamble, which is why we route that demand toward a campaign dosed over several months rather than a one-shot redirect. For teams that prefer to acquire placements one at a time on real publishers, buying a link directly from the publisher with no middleman sidesteps the expired-domain risk entirely.
Due diligence before you buy
The audit is the whole job. Start with the Wayback Machine and read the actual pages, not just confirm the domain existed. You are looking for a coherent topic, real editorial output, and the absence of the obvious red flags: pharma, casino, adult, or foreign-language spam injected at some point in the history. A domain that switched topics three times across its life has a confused link profile and almost no transferable trust.
Then audit the live link graph. Export referring domains, sort by whether they still resolve, and sample the actual linking pages. The questions are simple: are these links editorial or templated, do the linking sites still exist, and would a human editor have placed them. Anchor text distribution is the tell here. A natural profile is mostly branded and URL anchors with a long tail of topical phrases. A profile dominated by exact-match commercial anchors is a manipulation signature, and it will read that way to Google too.
Finally, price against liability, not against the metric. The seller is pricing the DR. You are pricing the DR minus the cost of the cleanup, the risk of the penalty, and the editorial work to rebuild credibly. This is the same discipline that separates an honest network from a metrics-arbitrage flip, and it is why we operate our own media in-house rather than reselling whatever drops cheap. If you would rather skip the acquisition entirely and place content on sites that already have standing, publishing a sponsored article on an established publisher gives you the placement without inheriting an unknown history.
What we see go wrong
The recurring failure is buying the number. An operator sees DR 55 for two hundred euros, skips the Wayback step, redirects it, and three months later the receiving site loses positions it held before. The expired domain did transfer something: a penalty. This is not rare, it is the default outcome when the only data point was a third-party metric.
The second mistake is ignoring topical drift. A perfectly clean expired domain from a regional news niche, repointed at a B2B SaaS money site, transfers almost nothing useful and signals manipulation by the mismatch alone. The links were earned in one context and are now being claimed in an incompatible one. Continuity is not a nice-to-have, it is the load-bearing condition.
The third is treating the rebuild as optional. Buying the domain is ten percent of the work. The other ninety is restoring content that justifies the existing links and gives the domain a reason to keep existing. Operators who skip this end up with a thin site that the 2024 policy describes almost word for word, and they are surprised when it deindexes. The honest version of this tactic is indistinguishable from running a real publication, which is the point.
Tactical takeaways
Read the history before the metric. The Wayback Machine and a live referring-domain crawl will tell you more in twenty minutes than any seller dashboard. Reject on liability fast and only spend the audit time on domains that survive the topical-continuity test. When you do buy, rebuild rather than redirect, and keep the niche close to what the domain already was. And size your expectation honestly: a well-chosen expired domain saves you the cold-start of a fresh registration, it does not hand you authority for free. The market that promises the latter is selling you the most inflatable number in SEO.