By Andrew Linksmith, Head of Link Acquisition · Update:
Link Exchanges: Where Google Draws the Line
A link exchange is an arrangement where two or more sites link to each other by agreement rather than editorial judgement. Google's spam policies name excessive reciprocal linking as a scheme — the operative word being excessive. Two businesses that genuinely work together linking to each other is normal; a systematic trading arrangement is not.
Natural mutual linking
Partnership links
Informal swaps
Systematic trading
Organised link circles
At one end of that range sit two integration partners linking because the integration exists. At the other sits a coordinated group exchanging links at volume across unrelated sites.
Almost every discussion of this topic collapses that spectrum into a yes-or-no question, which is why the advice is usually useless. The relevant question is not whether you have reciprocal links — every site does — but what proportion of your profile they represent and whether the underlying relationships are real.
Types of link exchange
Direct reciprocal
Site A ⇄ Site B
The simplest and most detectable. Two sites link to each other by agreement. Visible in any backlink tool in one click.
Three-way (ABC)
A → B
B → C
C → A
Adds a hop to obscure the reciprocity. The intent is concealment, which is itself the characteristic that makes the arrangement a scheme.
Four-way and larger circles
A → B → C → D → A
More participants, longer cycle. Harder to spot in isolation and easier to spot in aggregate, because the same domains keep appearing in each other's profiles.
Guest post swaps
A writes for B
B writes for A
Common and often not recognised as an exchange at all. If the arrangement is "I'll publish yours if you publish mine", it is a trade regardless of the content quality.
The three-way structure deserves comment because it is frequently described as the "safe" version. The added hop was designed to defeat a specific detection method — direct reciprocity checks — and that design intent is precisely what characterises it as manipulation. An arrangement built to be harder to see is not a compliant arrangement with better hygiene.
What Google's policy says
Google's spam policies list excessive link exchanges among the practices classed as link schemes, in the same section as buying and selling links and large-scale automated article campaigns. The full text is on Search Central and is worth reading in the original rather than in anyone's paraphrase, including mine.
Wording | What it does and does not cover |
"Excessive" | The qualifier that makes the policy workable. A single mutual link between two genuine partners is not what is being described. A profile where reciprocity is a defining characteristic is. |
"Link exchanges" | Covers direct swaps, multi-party circles and "link to me and I'll link to you" arrangements — regardless of how many hops are inserted. |
Partner pages | Not named as a violation. Linking to organisations you actually work with, on a page that exists for readers, is ordinary web practice. |
Enforcement is rarely a manual penalty. The common outcome is that the links stop counting — quietly, with no notification. That is worth understanding because it changes what "risk" means here: usually you lose the value of an arrangement you spent time on, rather than losing your site.
How exchange patterns get detected
Nothing about this is mysterious. Reciprocity is a property of a graph, and a link graph is exactly what a search engine maintains. You can measure your own exposure in about twenty minutes.
Reciprocity ratio
reciprocity = (domains that link to you AND that you link to) ÷ (total referring domains)
Export your referring domains.
Export your outbound linked domains.
Count the overlap. Divide.
Reciprocity ratio | Interpretation |
Under 5% | Unremarkable. Normal for any site with partners, suppliers and industry relationships. |
5–15% | Visible but usually explicable, particularly in ecosystems built on integrations and partnerships. |
15–30% | Reciprocity is becoming a defining feature of your profile. Worth understanding how it got there. |
Over 30% | The profile is substantially built on trading. This is what the policy language describes. |
Three other signals compound the ratio, and they matter more than the raw number:
Simultaneity. Links appearing in both directions within days of each other. Genuine relationships rarely produce perfectly synchronised linking; agreements do.
Anchor symmetry. Both parties using each other's target commercial phrase. Editorial writers describe; traders optimise, and they optimise in both directions.
Cluster repetition. The same set of domains appearing in each other's profiles across many combinations. One shared domain is coincidence; a recurring group of twelve is a structure. This is what catches multi-hop circles that survive a direct reciprocity check.
Method for running the full analysis is in how to read a backlink profile.
When link exchange is acceptable
There is a clean test, and it survives most edge cases.
Q1. Would this link exist if the other party had refused to link back?
Q2. Does the link help a reader of that page, independently of the arrangement?
Q3. Could you explain the relationship to a customer without it sounding like a technicality?
Three yeses: legitimate. Any no: you are running an exchange, whatever you call it.
Situation | Passes the test? |
Integration partners linking to each other's product pages | Yes — the integration exists regardless |
Supplier and distributor linking in both directions | Yes — a commercial relationship documented publicly |
Co-authored research where both parties link to the study | Yes — both contributed to the thing being linked |
Event organiser and sponsor linking mutually | Yes |
Two unrelated businesses agreeing to swap blog links | No — the link exists only because of the deal |
"I'll publish your guest post if you publish mine" | No — a trade with extra steps |
Joining a group chat that coordinates monthly link placements | No — this is the thing the policy names |
Note that the legitimate rows share a property: the link documents a relationship that produces something else. That is why they survive scrutiny — remove the SEO value entirely and the link still belongs on the page.
Safer alternatives to link swapping
Instead of… | Do this | Why it is safer |
Swapping blog links with a peer company | Co-produce research or a joint guide and both link to it | Both links point at a shared asset that genuinely exists — see linkable assets |
Trading guest posts | Pitch each publication independently on merit | Each editor makes a real decision; no reciprocal pattern forms |
Reciprocal "partners" pages between unrelated firms | Build integration or supplier pages only for relationships that exist | The relationship is verifiable outside the link |
Joining a link exchange group | Run reactive commentary and expert-source outreach | Editorial decisions, no reciprocity, no shared cluster — see digital PR |
Agreeing swaps to fill a monthly link quota | Reclaim unlinked mentions and claim supplier listings | Free, uncollected, and nobody has to link back — see 15 ways to get backlinks |
Every alternative above shares the same structural property: the other party's link is a consequence of something real rather than a condition of yours. That is the whole distinction, and it is more useful than any threshold.
Andrew Linksmith sells link placements and does not operate or broker link exchanges. Reciprocity thresholds reflect profile audits conducted by our team, 2023–2026, and are working heuristics rather than published limits. Google's position on link exchanges is set out in its spam policies.