Most link building advice fixates on volume. Build more links. Hit a quota. Watch the number climb. But anyone who has run campaigns across enough sites knows the uncomfortable truth: two links with identical metrics can produce wildly different outcomes, and the difference usually comes down to context.
A link is not a vote in isolation. It is a vote cast from a specific position, on a specific page, surrounded by specific words, on a site with a specific reason to be talking about your topic in the first place. Search engines have spent two decades getting better at reading that context. Link builders, by and large, have not caught up.
The page is the unit, not the domain
Domain-level authority metrics are convenient. They are also a blunt instrument.
When a strong domain publishes a page about garden furniture, the ambient authority of that domain does not automatically flow into a link on that page pointing at your fintech SaaS. The page carries its own topical signature, accumulated from its content, its internal links, and the external links pointing to it. That signature is what your link inherits.
This is why a link from a modest niche publication that has covered your subject for years often outperforms a link from a high-authority general news site that mentioned your industry once, in a listicle, three subheadings from where your link sits.
Practical implication: before evaluating a link opportunity, look at the page, not just the domain dashboard. What does this specific URL rank for? Does it have any external links pointing to it? Is it in the site’s internal linking structure, or is it an orphan sitting three clicks from anything that matters?
A page nobody links to and nobody navigates to is a page that passes almost nothing, regardless of the root domain’s numbers.
Topical proximity does the heavy lifting
Consider two links to a page about commercial insurance.
The first appears in an article about small business cash flow management, in a paragraph discussing how businesses budget for risk. The surrounding sentences discuss coverage, liability, and premiums. The link fits the argument being made.
The second appears in the same site’s article about office coffee machines, in a closing paragraph that pivots awkwardly to “and of course, don’t forget to protect your business.”
Same domain. Same page-level authority, hypothetically. Radically different signals.
The second link tells a search engine that this site places links opportunistically, without regard to whether they serve the reader. The first tells it that this site links the way editorial publications have always linked: because the reference genuinely advances the point.
The distance between your link and the nearest genuinely relevant sentence is one of the more useful heuristics available. If you have to write a bridging sentence to get to your anchor, the placement is probably wrong.
Position within the page carries weight
Not all placements on a page are equal, and this is not a controversial claim.
A link in the body of an article, embedded in prose, surrounded by relevant discussion, sits in the part of the page that search engines treat as primary content. A link in a sidebar, footer, or author bio sits in boilerplate — repeated across many pages, obviously not editorial, and discounted accordingly.
Within the body content itself, links appearing earlier in the article, in the sections where the core argument is being made, tend to be treated as more integral than links appended near the end. This is not a rule to game — stuffing links into your opening paragraph is transparent — but it does mean that a link naturally arising from the article’s central point is worth more than one tacked on as an afterthought.
The heuristic that has served me well: if you removed the link, would the sentence still make sense and still need a citation? If yes, the placement is sound. If the sentence collapses without the link, or if it would never have been written without the link in mind, the placement is artificial and reads that way.
The acquisition routes, honestly assessed
Understanding what makes a link valuable is one problem. Actually obtaining links in those positions is another. There are a limited number of routes, each with real tradeoffs.
Digital PR and original research. You produce something — a dataset, a survey, a genuinely novel analysis — that journalists and bloggers want to reference. Done well, this is the highest-quality acquisition method available, because the links arrive editorially, from pages written specifically about your subject, with anchor text you did not choose. Done badly, it is an expensive way to produce a report nobody reads. The failure mode is producing content that is interesting to you rather than interesting to the people you need to cover it. The cost is real, the timeline is long, and the results are unpredictable in a way that makes finance departments uncomfortable.
Guest posting. You write an article for a publication in your space, and the article contains a link to your site. When the publication is real, the editorial standards are real, and the article would have been commissioned regardless of the link, this is straightforward and legitimate — it is how trade publications have worked for a century. The problem is the enormous grey market of sites that exist primarily to sell guest posts, publishing hundreds of thin articles from anyone who pays. Those sites are identifiable, they are increasingly discounted, and links from them are at best neutral. Assessing the difference requires looking at whether the site has organic traffic, whether its articles are shared and cited, and whether it publishes anything at all that is not a guest contribution.
Resource and directory pages. Curated lists of tools, references, or organisations in a field. Genuinely useful ones exist and are worth pursuing. Most are abandoned, unmoderated, or link farms wearing a costume.
Niche edits. A link added to an existing published article, rather than a newly commissioned piece. The theory is sound: an established page that has already accumulated authority and topical relevance is, mechanically, a better host than a fresh page with neither. In practice the quality of the outcome depends almost entirely on the quality of the page and the honesty of the placement. Adding a genuinely useful reference to an article that would benefit from it — the way any editor updating an old post might — is defensible. Buying an insertion into an unrelated article on a site that sells placements in bulk is not, and the pages involved are usually identifiable as such. Anyone considering niche edits as part of a link programme should evaluate the target page on the same criteria as any other placement: does it rank, does it receive traffic, does it have its own inbound links, and does the insertion make the page better or merely longer?
Broken link replacement. Finding dead links on relevant pages and offering your resource as a replacement. Low volume, high acceptance rate when the replacement genuinely fits, and one of the few methods where the site owner receives clear value.
The pattern across all of these: the methods that work are the ones where the publisher has an independent reason to include your link. Where that reason is absent and has to be manufactured — with payment, with fabricated relevance, with a bridging sentence — the placement carries the marks of its manufacture.
What this changes about strategy
If context outweighs count, then the shape of a sensible link programme changes.
Fewer placements, evaluated more carefully. Ten links on pages that rank for terms adjacent to yours, embedded in paragraphs where they belong, will do more than a hundred links scattered across pages that have no topical relationship to your business and no traffic of their own.
Anchor text becomes less of a lever and more of a byproduct. When a link is genuinely editorial, the anchor is whatever phrase the writer naturally used. Distributions that look engineered generally are.
And evaluation shifts from spreadsheet metrics to page-level judgement. Does this page exist for readers? Would a link here be useful to someone reading it? Would you be comfortable if a competitor examined your backlink profile line by line and asked, of each placement, why it was there?
That last question is the one worth carrying into every acquisition decision. Not because search engines are watching in some paranoid sense, but because a link that cannot survive that question is a link that was never doing much work in the first place.
