How links still matter
Every few months somebody asks whether backlinks still matter now that the results page answers questions on its own. It’s a fair question, and it usually arrives with a suspicion attached: that the whole link business was always a bit of a racket.
Short version. They still matter. They matter less decisively than the industry selling them implies. And almost everything marketed as a shortcut is worse for you than doing nothing at all.
What a link is actually saying
A link from one site to another is a person deciding their readers should go and look at someone else’s page. That’s all it is. The reason it became the backbone of search ranking is that it was hard to fake at scale and it was a real signal of something real.
It carries more than one piece of information at once. There’s the endorsement itself. There’s the context, and context does a lot of work here: a link from a Danish trade body’s member list to a Danish supplier means something a link from a general directory never will. There’s the wording used in the link, which describes what the destination is about. And there’s the traffic, which everyone forgets, and which is the only part of a link that keeps its value no matter what any algorithm decides next.
That last point is the one I’d hold onto. If you evaluate a link by asking whether a real person might click it and be glad they did, you will almost never end up somewhere you regret.
Yes, but less than you’ll be told
Links have been the industry’s flagship product for as long as the industry has existed, and the reason is commercial rather than technical: you can invoice for a link. It arrives, you can point at it, it looks like a deliverable on a report. Very little else in this work has that shape, which is a fair part of why so much published advice overweights them.
Meanwhile the engines got better at the things links were standing in for. They can read a page and judge whether it’s genuinely about the subject. They can see whether a brand gets discussed, mentioned and searched for by name. A mention in a serious publication with no link at all now carries weight it wouldn’t have carried once.
Answer engines complicate it further. When a system builds a paragraph out of a few sources, the question isn’t purely which page has the most links pointing at it. It’s which page states the thing clearly and appears to belong to somebody credible. Links feed into that. They don’t settle it. There’s more on how that selection behaves in AI overviews, explained.
Where links still bite hardest is competitive commercial search. If five businesses all have decent pages about the same service, links are one of the few things left to separate them. But if you’re a small business in a narrow niche with three competitors, links are often not your constraint at all, and putting the budget there instead of into the pages themselves is a straightforward mistake.
What this looks like with no PR budget
Nobody running a small business is going to mount a campaign, and they shouldn’t try. Here’s what’s actually available, roughly in order of how often it gets ignored.
Relationships you already have. Suppliers, distributors, the manufacturers whose products you sell, the trade associations you already pay to belong to, the software vendors you’re a paying customer of. A surprising number of them run partner pages, member listings or customer directories. Most people never think to ask, and asking costs an email.
Things you already do locally. Sponsoring a club, hosting an evening, taking an apprentice, turning up to a business network. These produce links as a side effect whenever the organisation has a website, and the only work involved is asking whoever runs it to link the mention to you.
Writing where you genuinely know something. Not bulk guest posting, which is a different and much worse activity wearing the same clothes. One or two pieces a year, in a publication your customers actually read, on a subject where you have first-hand knowledge a general writer can’t fake, will do more for you than a pile of placements on sites nobody reads.
Being findable as a source. Journalists and trade writers are constantly looking for someone to quote. If your site makes clear who you are, what you do and how to reach you, and you reply while the story is still open, that’s most of the job.
Being worth linking to, which is the uncomfortable one. If there is nothing on your site a person would send to a colleague, no amount of outreach fixes that. A genuinely useful tool, a straight answer to the question everyone in your trade gets asked, a piece of information only you have because of how you work: that gets linked, and it keeps accumulating long after you’ve stopped thinking about it.
Expect all of this to be slow. A handful of real links in a year is a normal, healthy outcome for a small business, and it’s worth considerably more than a bulk package ever will be.
What I won’t sell you, and won’t recommend
Bought links. Paying for a link that passes ranking credit is against every major engine’s guidelines and has been for a very long time. That it’s common changes nothing about what happens when it’s identified.
Private blog networks. A stack of sites built for the sole purpose of linking to clients. They share hosting, registrars, templates and authors, and those shared patterns are precisely what gets a whole network removed at once, clients included.
Guaranteed placements. Nobody controls another publication’s editorial decisions, so a guarantee means one of two things: the placement is paid, or the site exists to sell placements. Same problem, different wrapping.
Bulk guest posting. Hundreds of near-identical articles placed on sites that accept anything, usually written by someone who has never done the job they’re describing. Readers notice. So do the engines, and so do the people you were hoping to impress.
Comment and forum drops. Nothing has come of this in a very long time, and it makes you look like a bot.
Link exchanges run at scale. Two businesses linking to each other because it genuinely helps their customers is fine and always was. A spreadsheet of sites swapping links in rotation is a scheme, and it reads as one from the outside.
Why the shortcut is a liability rather than a shortcut
The pitch never changes: faster, cheaper, everyone’s doing it. What it leaves out is who carries the risk, and the answer is you. The vendor takes the fee. The links attach to your domain. If those links get devalued, or the network they sit on is removed, the vendor refunds nothing and frequently isn’t reachable by then.
The failure modes are worth knowing before you decide. Links get quietly discounted, in which case you paid for nothing and nobody tells you. Or a manual action lands on the site, and you’re not ranking at all until the links are cleaned up and a reconsideration request has been read by a human. Or the network simply vanishes and your link profile changes shape overnight for reasons nobody in your business can explain to the board.
Cleanup is the expensive part, and it’s expensive in a way the original invoice never hints at. Working out which links are the problem. Getting them removed, where a real person answers an email, which is not most of the time. Disavowing the remainder carefully enough not to throw away the links that were helping. Then waiting, with nobody able to tell you when the traffic comes back.
The honest position is short: you don’t own a link you rented. The only thing you ever own is the reason somebody linked to you in the first place.
The questions this always raises
Someone emailed offering a package of links from high-authority sites for a flat fee. Ever worth it?
No. Read the email again: it’s the same email everyone in your trade received this week, the authority figure quoted comes from a third-party tool rather than from any search engine, and a site that sells placements has sold them to whoever else paid. What you’re buying is a footprint you can’t remove later.
A competitor is obviously buying links and doing well. How?
Sometimes it works for a while. That’s the uncomfortable answer and pretending otherwise would be dishonest. It works until it doesn’t, the timing isn’t predictable by anyone, and you’d be taking on a risk you have no way to price.
Should we disavow links we bought years ago?
Possibly, and carefully. The disavow file is easy to misuse and removing links that were quietly helping does real damage. It’s worth going through the profile properly rather than pasting in a domain list and hoping.
Do we need links to be cited in AI answers?
Not directly, as far as anyone outside those companies can tell. But the sources these systems lean on tend to be the ones already treated as credible elsewhere, and links are part of how that credibility gets established. AI overviews, explained covers the rest of it.
Why isn’t there a free tool for this, like the others?
Because there’s nothing honest to automate. A link is worth something because of who published it and why, and no browser-side tool can assess that from text you paste in. A tool that scored links anyway would just be inventing a number. Links & Authority is judgement work for exactly that reason, which is also why it’s the slowest thing here.
If you’d rather not work this out alone
Slow, manual, judgement-led work on the relationships and pages you already have. No packages, no placements bought on your behalf, and no promise about where anything ends up ranking.