What makes a backlink high quality, and how to judge one quickly

Most advice about link quality is written for people browsing a menu. You are not browsing. You have one thing in front of you — an email offering a slot, a directory that wants £90, a supplier who says they will add you to their partners page — and you need to decide whether to spend the next four hours on it or delete it.

That is a different question, and it has a faster answer. What follows is a test you can run in about a minute, in a browser tab, without a subscription to anything.

The six questions

Ask them in this order. Most bad opportunities fail on the first two and you can stop.

1. Is the linking page about roughly the same thing as the page it points to?

Not the site — the page. A general home-improvement blog might be broadly relevant to a plumber, but the specific article the link would sit in is what a search engine reads and what a reader is actually looking at. "Fifteen kitchen colour trends" is not relevant to emergency drain clearance, whatever the domain says.

2. Would a real person plausibly click it?

This is the single most useful question, because it is nearly impossible to fake and it correlates with almost everything else. Read the sentence the link would sit in. If a person reading that sentence, for their own reasons, would want to go where the link goes, you have something. If the link exists only because you paid for it to exist, the person reading has no reason to touch it — and that is usually visible.

3. Was it placed editorially, or did money or automation place it?

Editorial means someone with an audience decided your page was worth pointing at. Google's spam policies treat links intended to manipulate ranking — including links bought or exchanged for that purpose — as link spam, and Google's documentation asks that paid or sponsored links be qualified with rel="sponsored" or rel="nofollow". A qualified paid link is not a crime; it is just not a ranking signal. Decide which one you are buying before you pay.

4. Where does it sit on the page?

Inside a paragraph of body copy, surrounded by relevant sentences, is the strong position. A footer, a sidebar block, a "resources" wall of two hundred outbound links, or a chunk of boilerplate repeated site-wide is the weak one. If the same link block appears on every page of the site, you have one link, not four thousand.

5. Does the site show signs of existing to sell links?

There is a recognisable shape. Categories that have no relationship to each other — finance, dentistry, casinos, CBD, removals. Posts with a single commercial outbound link in the second paragraph and nothing else of substance. A "write for us" page that quietly turns into a rate card. Author names with no presence anywhere else. No comments, no newsletter, no reason for anyone to return. One of these means nothing. Four of them together means the site's product is links, and its readers are other link buyers.

6. Does the page itself have any standing?

Does anything link to *it*? Does it rank for anything? Does it get shared? A page nobody has ever visited on a well-known domain passes very little along. Domain-level reputation is real but page-level reality matters more than most people assume.

A one-minute scoring pass

| Question | Strong answer | Weak answer | |---|---|---| | Topical fit of the page | Same subject, same intent | Same broad industry at best | | Would a person click | Yes, for their own reasons | Only because it is there | | How it got placed | Someone chose to add it | Paid, swapped, or automated | | Position on the page | In body copy, mid-article | Footer, sidebar, link wall | | Site's apparent purpose | Serves an actual audience | Publishes to sell placements | | The page's own standing | Ranks, gets linked, gets read | Orphaned, no visible readers |

Four or more strong answers and it is probably worth real effort. Two or three, and it is worth effort only if the effort is small. Zero or one, and the honest answer is that you are buying a decoration. If the mechanics behind why some of these matter and others do not are hazy, the explanation of how backlinks actually work covers the underlying model, and the glossary definition of a backlink is the sixty-second version.

About authority scores

Every link-buying conversation eventually produces a number: DR, DA, TF, AS. These are inventions of the companies that sell the tools. Each vendor built its own model on its own crawl of the web, and each publishes a different number for the same site. Google does not publish a per-site authority score. There is no public figure to check them against.

That does not make them useless. As a rough comparator — is this site in a completely different league from that one — they are fine, and they are much faster than manual inspection. What they cannot do is tell you whether a link will help *you*, because they know nothing about relevance, placement, or whether a human will ever see it.

The practical failure mode is that these scores are the one thing in the whole equation that can be manufactured. A site can be built to score well and be worth nothing. Nobody has ever built a site that a real audience genuinely reads and links to in order to inflate a third-party metric. Treat the score as the last question you ask, not the first.

Four worked judgements

The businesses below are illustrative — invented to show the reasoning, not case studies of real clients or measured outcomes.

Ferndale Boiler Care, offered a paid slot in a "Top 50 home services blogs" roundup

An email arrives: £180 for inclusion in a roundup post on a site with a DR in the sixties. Ferndale is a two-person heating firm covering one town.

Relevance is weak — the roundup is national and generic. Clickability is near zero; nobody reads a top-fifty list and rings the twenty-third entry. It is plainly paid. Position will be a list item among forty-nine competitors. The site's categories run from home services to personal finance to travel. The score is the only strong thing about it.

Verdict: decline. One strong answer out of six, and the one that is strong is the manufacturable one. The £180 buys a line of text on a page nobody reads.

Ferndale Boiler Care, sponsoring the local rugby club

£120 a season, name and link on the club's sponsors page.

Topical relevance is poor — a rugby club page is not about boilers. The link sits on a sponsors wall, which is the weak position. It is paid, and it should be marked as sponsored. But: local people read that page, some of them own boilers, and the club will mention the sponsorship in its newsletter and at the ground.

Verdict: yes, but not as a link. This is local sponsorship that happens to include a link. Judged as a ranking asset it fails; judged as £120 of community visibility in the one town Ferndale serves, it is reasonable. The distinction matters because it changes what you expect from it and how you measure it. That is the same reasoning applied in the honest cost comparison of paid versus organic visibility.

Larkspur Invoicing, invited to write a guide for a trade association's resource library

A small accounting-software firm is asked by an industry association to contribute a practical guide on invoicing rules for its members' resource section. No money changes hands. The guide will take two working days.

Relevance is exact. Members would click it because they need the information. It is editorial — the association chose to ask. The link will sit in the author credit and possibly in body copy. The association exists to serve members, not to sell placements. The resource library is one of the more-visited parts of its site.

Verdict: strong yes, and it is worth the two days. Six for six, and the by-product — being the firm the association asks — is worth more than the link. Opportunities like this are the reason getting your first backlinks when nobody has heard of you is mostly a relationships problem rather than a technical one.

Larkspur Invoicing, unsolicited offer on a "business news" site

£220, permanent placement, DR 72, "guaranteed dofollow, no sponsored tag".

The last clause is the tell. The seller is explicitly offering to breach Google's guidance on qualifying paid links, which means the site is either doing this at scale or willing to. Check the site: two hundred posts in six months, every one with a commercial outbound link, categories spanning crypto to hair transplants.

Verdict: decline, and do not reply. The high score is exactly what is being sold, and it is the only thing on offer. If you are weighing this class of purchase seriously, the honest cost comparison of buying backlinks lays out what the money actually buys and what the exposure is.

The cost side, which most guides skip

A link's quality is only half the decision. The other half is what it costs you in hours. A perfect editorial link that requires two days of writing has to be worth two days. A mediocre but genuine listing that takes eleven minutes to submit has a very low bar to clear.

Rank your opportunities by quality *divided by* effort, not by quality alone. This usually reorders the list dramatically, and it is why the boring wins — supplier directories, trade bodies you already pay to belong to, the local chamber, the software you already use and its customer list — tend to beat the exciting ones.

Two mistakes that survive most advice

Treating nofollow as worthless. A qualified link that sends you actual visitors is a marketing asset. It is just not a ranking asset. Judge it on the traffic, and be honest about which of the two you were buying.

Assuming you need to clean up links you did not build. Google's documentation is clear that it is generally able to ignore links it does not trust, and the disavow tool is aimed at cases where you know you have a manual action or a history of buying links. For most sites, the odd spam link appearing in a report is not a problem worth solving.

Where this framework stops being useful

Sometimes you run the six questions across every opportunity available to you and the honest total is: there aren't enough of them, and the good ones will take a year. That is a genuine outcome, not a failure of the method, and it is worth reading alongside a realistic timeline for ranking a brand-new website before deciding you have done something wrong.

At that point some businesses look at paid placement instead — including this marketplace, where position is a published function of verified spend and every figure is visible on the public rankings and the spend behind them and explained in full on the transparency page. Be clear-eyed about what that is: a paid listing is a paid listing. It should be marked as such, it does not pass ranking signals, and it will not do anything for your organic search position. It buys visibility on one page, on stated terms, for a stated price.

If what you actually need is to rank in Google, no amount of paid placement anywhere substitutes for the slow work described above. Run the six questions, take the cheap wins, write the association guide, and let the domain age.

Short FAQ

Is a DR 80 link always better than a DR 30 link? No. A DR 30 industry publication whose readers are your customers beats a DR 80 general site whose readers are not, on every question except the score itself.

How many links do I need? There is no number, and anyone quoting one is guessing. Competitiveness of the query and the strength of everything else on your site decide it. The plain explanation of how Google ranking works covers why a target count is the wrong frame.

Can I check any of this without paid tools? Mostly, yes. Read the page. Read three other pages on the site. Look at whether anyone comments, subscribes, or shares. Search a distinctive sentence from one of its posts to see whether it is original. That is four minutes and it answers five of the six questions.

What if the offer is free but the site looks like a link farm? Free does not change the analysis. The cost was never the problem — the placement was.