Introduction
Outreach link building starts from a fact about the web: the opportunity already exists on somebody else's page. Someone maintains a resource list that should include you and does not. Someone links to a page that died three years ago. Someone wrote about your product by name and did not link to it. Someone lists four of your competitors in a roundup. In every case the page is already published, the editorial decision to cover the topic has already been made, and the only thing missing is that nobody has told the owner.
Your job is to find those situations, qualify them, and ask. That is the whole category. Thirteen tactics sit here, and the differences between them are mostly differences in how the opportunity is found, not in what you do once you have found it.
What separates outreach from the other two categories
In the earned category you make something and wait. In the placed category you pay and the link appears. Outreach is the middle: you do research to find a real opportunity, then you persuade a human being to act on it. Both halves matter, and most programs are bad at one of them.
The research half is a skill that transfers across every tactic here. Search operators, backlink exports, crawling a competitor's referring domains, checking whether a page is indexed and gets traffic, finding the person who actually controls the page rather than the address on the contact form. Get good at that and the specific tactic becomes almost interchangeable — broken link building and resource page link building are the same workflow pointed at different queries.
The persuasion half is where programs fail. You are asking a stranger to edit a page they may not have touched in two years, for the benefit of somebody they have never heard of, at no gain to themselves. If your email does not make the value to them obvious in the first two lines, the rest of it does not get read.
Response rates are low, and the arithmetic follows from that
I am not going to quote you an industry response rate, because the published figures vary by an order of magnitude depending on who is doing the publishing and what they are selling. What is not in dispute is the direction: cold outreach asking a stranger for a link gets a small single-digit percentage reply rate, and the proportion of replies that end in a live link is smaller again. Assume most of what you send goes unanswered forever, and build the program around that assumption rather than being disappointed by it.
Three consequences follow, and they are the whole strategy of this category.
Volume is structural, not optional. If a small percentage of qualified prospects convert, a campaign built on twenty prospects is not a campaign, it is a coin flip. You need enough prospects that the conversion rate can express itself. This is uncomfortable for people who want link building to feel artisanal, but the arithmetic does not care.
Volume without qualification is just spam, and it degrades. The obvious response to a low conversion rate is to send more emails, and if you do that by loosening your standards you get worse results, not more of the same results. Unqualified prospects reply less, the sites that do reply are worse, and your sending domain's reputation takes damage that affects every campaign after it. The correct response to a low conversion rate is more qualified prospects, which means more research, which is slower and more expensive. Anyone promising outreach volume without a proportionate research cost is either using somebody else's list or not qualifying at all.
Per-email effort has a ceiling and a floor. A fully handwritten email to every prospect does not scale and usually is not necessary. A template with the site name merged in gets deleted. The workable middle is segmenting prospects into groups that genuinely share a reason to link, writing a real message per segment, and personalizing the specific detail that proves you looked at the page. Two sentences of genuine specificity beat five paragraphs of flattery.
The tactics in this category
Six are graded safe: broken link building, resource page link building, local link building, unlinked mention reclamation, link reclamation and competitor link gap analysis, plus international link building. What they have in common is that the ask is legitimate on its face. You are telling somebody their page has a dead link, or that a resource they curate is missing an entry, or that they mentioned you without linking. If they say no, nothing bad happens and nothing needs explaining.
Four are situational: guest posting, directory and citation links, image attribution and photo credit, and link roundups. Each one is fine at small scale and executed with judgment, and each one has an industrialized version that is not fine. Guest posting is the clearest case — a genuine contributed article to a publication that would have commissioned it anyway is editorial, and buying placements on a network of sites that exist to sell them is not, and the same two words describe both.
Two are graded high risk: link exchanges and reciprocal links, and .edu and scholarship links. Both are technically outreach — you find a page and you ask — but the ask itself is the problem. In one you are trading links, which is named explicitly in Google's link spam policies. In the other, the scholarship is a pretext for the link and the pattern is well known to everyone who looks for it.
What outreach cannot do
It cannot reach the top of the quality distribution. On my own site's data, only 446 referring domains out of 22,260 sit at Trust Flow 41 or above — the top 2% — and essentially none of those came from anyone asking. Publications with real editorial authority do not add links because a stranger emailed to point out a broken one. They link because their writer found your work while researching a story. That is the earned category, and no amount of outreach volume substitutes for it.
Outreach also cannot fix a site that has nothing worth linking to. Every tactic here ends with somebody looking at your page and deciding whether to point their readers at it. If the page is thin, or is a product page with no information on it, the answer is no regardless of how good your email was. I have watched a great deal of outreach effort wasted on sites whose real problem was that they had not published anything worth citing. Fix that first; the outreach gets dramatically easier afterward.
What outreach is genuinely good at is the middle of the distribution: relevant, real, moderately trusted sites in your subject area, acquired steadily and predictably by people who are willing to do the research. That is most of what most sites need, and it is available to anyone prepared to do the work.