Introduction
An earned link is one you did not ask for. You published something, built something, said something, or did something in public, and a third party decided independently that it was worth citing. You had no contact with them. You did not negotiate. You often do not find out until the link shows up in a backlink export weeks later.
That definition is narrow on purpose. Plenty of link building calls itself "earned" because it feels virtuous. If you emailed someone and asked, that is outreach, and outreach is a perfectly respectable category with its own section. Earned means the decision to link was made by somebody else, for their own reasons, without you in the room.
Why it is the hardest category
Every other category has a lever you can pull. In outreach you can send more emails. In the placed category you can spend more money. In earned link building the only lever is the quality and visibility of the thing itself, and quality is slow, expensive and unreliable in a way that neither of the other two is.
You also cannot forecast it honestly. An outreach campaign has a response rate you can estimate from your own history within a factor of two. A digital PR campaign either lands or it does not, and the difference between a study that gets picked up by forty publications and one that gets picked up by none is often not the quality of the work — it is what else happened in the news cycle that week. I have watched genuinely good research get buried because a bigger story broke on the day it went out. Anyone who tells you they can forecast earned coverage precisely is selling you something.
And it is slow at the front. The first three months of an earned program usually produce nothing you can point at. There is a research phase, a production phase and a distribution phase before a single link exists. That is a hard sell internally, which is exactly why so many programs abandon earned tactics after one quarter and go buy something instead.
Why it has the highest ceiling
Here is the number from my own site's data that changed how I think about this. Across 1,301,839 links from 22,260 referring domains — a profile where no link was ever bought — only 446 referring domains are at Trust Flow 41 or above. Trust Flow is Majestic's measure of how trustworthy the sites linking to a page are, on a 0 to 100 scale. Those 446 domains are 2.0% of the profile, and they accumulated over roughly twenty years. That is about twenty-two genuinely strong referring domains a year, from two decades of writing, speaking and being quoted.
You cannot buy your way to that number and you cannot email your way to it either. Publications that carry real trust do not respond to outreach templates and do not sell links in the first place. The only route to the top of the quality distribution is to be worth citing, which is what this category is.
The survival data points the same way. On my profile the median referring domain stops linking after about 1,080 days — call it three years — and 49.6% of all links ever recorded have been lost. But survival scales sharply with trust: domains at Trust Flow 0 have a median lifespan of 859 days, while domains at Trust Flow 61 and above run to a median of 3,353 days. Roughly four times longer. An earned link from a strong source is not just better on the day you get it; it is still there when three cohorts of cheaper links have quietly disappeared.
One more finding worth carrying into this category: 97.4% of my lost links were lost while the source page was still perfectly reachable. Only four lost links out of 645,202 went to a 404. Links do not usually die because sites die. They die because somebody redesigned a page, rewrote an article, migrated a CMS, or pruned an old post — and an editorially motivated citation inside a substantive article survives all four of those events far better than a link that was inserted into a page nobody has a reason to keep.
What the twelve tactics here have in common
They look different on the surface. A data study, a podcast appearance, a conference sponsorship and a testimonial are not obviously the same activity. But every tactic in this category shares four properties.
- You produce something before anyone links to anything. There is an asset — research, a tool, a talk, an event, a quotable opinion. The asset exists first and the links are a downstream consequence. This is the opposite of outreach, where the opportunity exists first and you go find it.
- The decision belongs to someone else. A journalist, an editor, a blogger or a program organizer decides. Your influence is upstream: make the thing genuinely useful, make it easy to cite, make it easy to find, and be reachable.
- The output is unpredictable per unit and stable in aggregate. Any single campaign may produce nothing. A sustained program across a year produces a distribution you can plan around. This is why earned work only makes sense as a program, never as a one-off campaign — and why it is so often abandoned by organizations that fund it as a one-off campaign.
- The links that result are the durable ones. Editorial citations inside real articles, on sites with their own audiences, are the links that survive redesigns and outlive the person who acquired them.
Eight of the twelve are graded safe. Four are situational, and it is worth understanding why, because it is the same reason in each case: the tactic starts as genuine earned activity and degrades into something else when it is scaled. Sponsorship becomes a paid link when the only reason for the sponsorship is the link. Testimonials become a link scheme when you write forty of them for products you have never used. The skyscraper technique becomes ordinary outreach spam when the "better resource" is not actually better. Infographics were an earned tactic that got industrialized into a placement tactic. In every case the grade reflects what happens at volume, not what happens when a competent person does it once, properly.
Who this category is realistic for
Earned link building needs three things: something to say, somebody able to produce it to a publishable standard, and enough patience to fund a program for longer than a quarter. If you have proprietary data, unusual operational experience, or a genuine expert who will put their name to an opinion, you can start today. If you have none of those and no budget to create them, the honest answer is that outreach will serve you better for now, and you should build toward this category rather than pretend you are already in it.
The failure mode I see most often is an organization that wants earned results on an outreach timeline. They commission a study, give it three weeks, get four links, and conclude that digital PR does not work. What did not work was the schedule.