Backlinks: What They Are and What Makes One Worth Having

Quick answer: Backlinks are links from other websites to yours. Each one carries three things: a path a crawler can follow, an association created by its anchor text and surrounding context, and a share of the linking page’s own standing. Relevance and editorial intent decide the value, and the first link from any domain does most of the work.

This is the foundation piece for everything else we publish on off-page work, so it sets definitions rather than tactics. If you want prospecting workflows, tool comparisons, agency delivery models or the relationship between off-page and on-page work, those are separate subjects with their own answers, and this post deliberately stays on the fundamentals: what a link is mechanically, what makes one worth having, how they influence rankings in a system that now has a great many other inputs, and what a realistic profile looks like when you are operating in a market the size of Singapore. That last point is the one most guidance gets wrong for a local reader, because almost all of it is written from a market with a hundred times as many available linking domains. Our SEO services page covers where this sits commercially.

The Mechanics: What a Link Actually Is

A backlink is an HTML anchor element on someone else’s page whose destination is a page on yours. Mechanically it is unremarkable. What matters is the three distinct things it carries.

A crawl path. Search engine crawlers discover pages by following links. A page nobody links to, internally or externally, is hard to find and slow to be indexed. This is the oldest and least glamorous function of a link and it still matters most on new sites.

An association. The anchor text, the sentence around it and the topic of the linking page together tell a search engine what your page is about, in someone else’s words. Third-party description carries weight precisely because you did not write it. This is why a link from a page about corporate tax filing means something different from an identical link on a page about office furniture.

A share of standing. The linking page has its own accumulated signals, and linking passes some of that along. How much, and how it is calculated, is not public and has changed considerably. That it happens at all is the reason the entire industry exists.

The rel attribute modifies the third of these, not the first two. A link marked nofollow, sponsored or ugc is, since 2019, treated by Google as a hint rather than a strict instruction, which means such links may still be used. More practically: a nofollow link from a publication your customers read still sends people, still creates brand searches and still creates the association. We would not sort a prospect list by rel attribute, and the habit of doing so has probably cost local businesses more good opportunities than any algorithm update.

What Makes One Worth Having

Six factors, roughly in order of how much they matter in practice.

Relevance. Topical proximity between the linking page and yours. A link from a Singapore logistics trade publication to a freight forwarder is worth a great deal more than one from a general-interest blog with a much larger audience, because relevance is what makes the association meaningful.

Editorial intent. Was the link given because someone thought the page was worth citing, or was it placed because a mechanism placed it? Directory submissions, profile pages and signature links sit at the low end. A journalist citing your data sits at the high end. Search engines have spent fifteen years getting better at telling these apart.

Position and context on the page. A link inside the body of a piece, surrounded by relevant text, is worth more than one in a footer, sidebar or a list of two hundred partners. Body placement also correlates with the link being clicked, which is the honest test.

Whether anybody links to the linking page. A page that nothing points at has little standing to share. This is the grain of truth inside the authority metrics, and it is why a link from the main page of a small association can outperform one from an orphaned page on a large site.

Whether the domain is new to you. This is the single most under-appreciated factor. The first link from a domain does most of the work available from that domain, and subsequent ones return sharply less. Thirty links from one partner site is not thirty links; it is closer to one. Count domains, always.

Plausible human traffic. If nobody could realistically click it, it is a notional asset. Links that send real people tend also to be the links that rank well, because the same properties cause both.

The pattern in that table is the whole lesson. The valuable rows all involve something real happening in the business. The cheap rows are cheap because anyone can have them, and anything anyone can have carries almost no information.

How They Influence Rankings Now

Honesty about this has improved in the industry, but the popular account is still stuck around 2012.

Links remain a ranking input. Google has confirmed as much repeatedly, and no serious practitioner thinks otherwise. They are not the only input and their relative weight has declined as the systems have acquired more signals to work with.

Most links are filtered before they count. A great deal of what a third-party index shows you is discounted: scraper copies, auto-generated directories, known networks, links from pages with no standing, and patterns that look manufactured. The effective profile is much smaller than the reported one, which is one reason reported totals are such poor guides.

Relevance is evaluated at query level. The same link can help for one query and be immaterial for another, because what is being assessed is the relationship between a query, a page and its context. A general authority number cannot represent that, which is the structural reason those metrics mislead.

Volume has stopped being the lever it was. Adding domains helps most when you have very few, and the return flattens quickly after that. Going from 5 domains to 25 can change outcomes materially. Going from 100 to 400 in a local niche usually changes nothing you can measure, and the effort would have been better spent elsewhere.

Low-competition local queries often need very few. This is the most useful and least repeated fact for a Singapore SME. Plenty of page-one results for local service queries sit on sites with a handful of referring domains, because relevance, clear local signals and a crawlable site are carrying the result. We have found, more often than not, that a local site’s constraint was not its link profile at all, which is the first thing our SEO audit and consulting work is designed to settle before anyone is sold outreach.

What a Realistic Singapore Link Profile Looks Like

Here is the structural point that changes everything for a local reader.

The supply of linking domains in Singapore is finite and small. Write out the honest sources: the active national and business news outlets, a thin layer of trade and sector publications, the statutory boards and government agencies, a few dozen tertiary institutions, the professional bodies and trade associations, the chambers of commerce, a modest set of genuinely curated local directories, the events and awards organisers, and the commercial web of suppliers, distributors, stockists, partners and clients. That is close to the complete list, and you could draft most of it in an afternoon. There is no long tail of thousands of independent niche blogs the way there is in a large English-language market.

So the numbers that constitute “good” are much lower than published benchmarks. In the local accounts we have audited, sector leaders in professional services, trades and specialist retail commonly sit in the tens rather than the hundreds of referring domains. A guide telling a Singapore accountancy practice that it needs 300 referring domains to compete is applying a benchmark from a market with a different supply curve.

It also means heavy overlap. Run the profiles of four competitors in the same local niche and the same association, the same two publications and the same three directories recur. That is not a sign of coordination. It is what happens when everyone draws from the same small pool. The practical effect is that the reachable gap between you and the leader is usually small in absolute terms and closable with unglamorous work.

And it means saturation. Once you hold the association listing, the trade publication mention, the supplier pages and the handful of curated directories, competitor-derived prospecting runs out. After that the work is editorial: giving local publications, associations and partners a reason to mention you. Businesses that do not make that transition tend to keep buying volume, and volume in a saturated pool means descending into sources that are worth nothing or worse.

Those ranges are patterns we have observed rather than published benchmarks, and they vary. The reason to state them anyway is that the alternative, importing a number from a large market, is reliably worse. Sector differences are sharp enough that we scope link expectations per vertical rather than as a flat target, which is the logic behind our industry SEO work.

Two sectors illustrate how wide the variation gets. Property is dominated by a small number of portals that hold most of the citable ground, so an individual agency is competing for a thin remainder, which is the situation our real estate SEO page addresses.

Education sits at the opposite end, because institutions, ministries, agencies and parent media all link outward as a matter of ordinary practice, so the available pool is larger and the realistic targets are higher. That difference is the reason our education SEO page sets different expectations.

Where Local Links Genuinely Come From

Practical, in rough order of return for effort.

Things you have already earned and never claimed. Suppliers, distributors, brands you stock, software you are certified in, clients who would gladly name you, professional bodies you already pay. On most accounts we review, several of these exist and are unlinked. This is the cheapest work in the entire discipline.

Association and professional body membership. You are paying the fee anyway. Complete the profile, include the link, and check annually that a site redesign has not quietly removed it.

Trade and sector publications. Small readerships, exactly the right readers. They need explanatory material, data and comment, and they are reachable by a direct email in a way national press is not.

Local business and national press, via something genuinely reportable. Your own data on a local question, a clear position on a change in regulation, or a story about the business that is actually a story. The media pool here is small enough that a relationship with three journalists is a meaningful asset.

Community, events and sponsorship. Organiser sites list sponsors and speakers. Choose things you would do regardless, and the link is a by-product rather than the point.

Tertiary institutions. Guest lectures, internship programmes, industry-panel participation and research input all produce pages that link out. Slow, high value, and durable.

Client and partner case studies on their sites. Under-asked. When a project has gone well, the other party is often pleased to write about it, and in agent-led and referral-led businesses that single habit can account for a large share of the profile. Our property agent results write-up shows how an agent with a strong referral network built her own search presence alongside it.

Unlinked mentions. You are already being named somewhere without a link. Finding those and asking is high-conversion because the editorial decision has already been made in your favour. Where a business has outlets, this layer interacts heavily with listings work, which is the subject of our local SEO in Singapore page.

What to Avoid, and How to Recognise It

Buying links breaches Google’s link spam policies. The reason to describe the schemes is so you can recognise them when they are offered to you, which they will be.

Fixed monthly link quotas. A promise of a set number of links per month regardless of what is happening in the world, at a set fee. Nobody can control how many editorial decisions other people make, so a guaranteed number implies a controlled supply, and a controlled supply implies purchased placements.

Private blog networks. A stable of sites built to host links, dressed as a publisher portfolio. Tells include refusal to name the sites before payment, thin content, and no real audience.

Marketplaces selling placements. Menus of sites with prices attached to a metric. The transaction is the link, which is precisely what the policies address.

Paid “guest posting”. Contributing genuinely useful writing to a publication is legitimate and always has been. Paying a fee for a placement is not, and the distinction is whether money changed hands for the link.

Sitewide footer link exchanges. Low value even setting policy aside, because of domain-level diminishing returns.

Comment, forum and profile spam. Ineffective for a decade and it makes the profile look manufactured.

Anchor text over-optimisation. Even with legitimate links, a profile where a large share of anchors are the same commercial phrase looks engineered. Natural profiles are dominated by brand names, URLs and incidental phrases. If you are in a position to influence anchor text, ask for the brand or a natural phrase, not the money keyword.

The defensive posture is simple. Do not buy links, and do not hire anyone who does, because the liability sits with your domain and not with the vendor. Ask any provider to name the specific sites and show you the outreach; an inability or refusal to do so answers the question. Our about page sets out how we think about this rather than a rate card.

Field notes: Our B2B ecommerce case study shows what a modest, legitimate link profile built from a business’s own industry looks like. The wholesale kitchenware supplier in Ubi had operated for eleven years through trade show relationships and distributor networks, yet its Domain Authority was 11 and its organic enquiries came only from brand-name searches by existing contacts. From Month 4 to Month 8, it was submitted to the Singapore Business Federation directory, the Enterprise Singapore supplier registry, HardwareZone Business and 12 industry-specific trade directories covering F&B, hospitality and food manufacturing. Every one of those is a source a trade buyer might actually use, and none of them required anything other than being a legitimate supplier in that industry. By Month 8, Domain Authority had reached 22 and monthly organic trade enquiries had moved from 5 to 28, alongside 14 category pages, 6 industry vertical pages and trade buyer content, so the results belong to the whole programme. The relevant sources in most Singapore sectors are a finite, nameable list, and working through it beats chasing global benchmarks.

Our Take

Backlinks are a genuine ranking input and a poor organising principle for a small business’s marketing. The distinction matters because the industry sells the second while the evidence supports the first.

The trade-off worth resolving: link work is slow, uncertain and hard to attribute, while on-page and technical work is fast, controllable and measurable. That does not make link work optional, because in competitive local queries you will eventually need the profile. It does mean the order matters. Establish whether links are actually your constraint by counting what the sites above you have. If they have fewer than you, the answer is somewhere else and no amount of outreach will find it.

Conventional wisdom in this field measures links by volume. In a market with a finite and countable supply of linking domains, volume is close to the wrong dial entirely. Ten relevant, editorially given links from sources your buyers actually read will outperform two hundred from anywhere, and past a certain point volume is only achievable by descending into sources that carry risk rather than value. Count domains, name your sources, reclaim what you have already earned first, and treat the association listing and the trade publication mention as the serious assets they are here.

Where the real constraint turns out to be page-level rather than profile-level, that is technical SEO territory, and it is a faster and more controllable piece of work than any outreach programme.

Expectations differ again in hospitality, where the aggregators and booking platforms hold most of the citable ground and a property has limited ability to change that. Our boutique hotel results write-up is organised around working within it.

What a programme should cost to run properly, and where the diagnostic ends and the ongoing work begins, is set out on our pricing page.

We count referring domains, not total links, for every client audit, because in our experience that single change in what gets measured prevents most of the bad decisions this topic invites. Our team has watched businesses spend a year chasing link volume while the sites actually outranking them had a fraction of the count, and the eventual fix was always the same: fewer, better, more relevant sources. We recommend running that domain count before spending a dollar on outreach.

Frequently Asked Questions

How many backlinks do I need to rank in Singapore?

There is no universal number and the useful answer is empirical. Take the three queries that matter most, look at who occupies the top three results, and count their referring domains. In many local service niches the answer is surprisingly low, and it is not unusual for page-one results to sit under 20. Aim to be credible within the range you observe rather than at a benchmark imported from a market with a much larger supply of linking sites.

Is it links or domains that count?

Domains, decisively. The first link from a site delivers most of what that site can offer and additional links from the same domain return sharply less. This is why a partner putting your logo in a footer across four thousand pages is worth roughly one link, and why any report leading with total backlinks is showing you an inflated figure. Count and report referring domains.

Do nofollow links help?

They are not worthless. Since 2019 Google treats rel attributes as hints rather than absolute instructions, so such links may still be used. Independently of that, a nofollow link from a publication your market reads sends real people, generates brand searches and creates a public association with your business. The practical advice is to stop filtering prospects by rel attribute and start filtering by whether a relevant human would read the page.

Can I buy backlinks?

Buying links to influence rankings breaches Google’s link spam policies, and the risk sits with your domain rather than with whoever sold them. Beyond the policy question, it is a poor commercial decision in a small market: the purchasable supply is low quality, the same networks serve your competitors, and money spent on an association membership, a sponsorship or a piece of content a trade publication actually wants goes further. Decline it, and decline providers who offer it.

What is a good anchor text distribution?

A natural one, dominated by your brand name, bare URLs, and incidental phrases such as the name of the article or a generic phrase like the company’s service in someone else’s words. Exact commercial phrases should be a small minority. If a large share of your anchors are the same money keyword, the profile looks engineered regardless of how the links were obtained. Where you have influence, ask for the brand name.

How long does it take for a new link to have an effect?

Weeks at the fast end and often much longer, and the effect is usually not attributable to that single link. The crawler has to fetch the page, the link has to be processed and assessed, and any ranking movement then competes with everything else changing at the same time. In practice a link-building programme is judged over two to three quarters on direction, not on individual placements, which is why proving the work requires recording acquisition against attrition from the start.

Do I need to worry about bad links pointing at my site?

In the overwhelming majority of cases, no. Every site of any age accumulates scraper copies and auto-generated directory entries, and search engines discount them as a matter of routine. The situations that deserve attention are a sudden large volume appearing in a short period, which often indicates a compromised site somewhere, and a documented history of purchased links, which is a cleanup project. Otherwise, doing nothing is the correct action.

Should I use the disavow tool?

Rarely, and not as a maintenance habit. It is a blunt instrument, easy to misapply, and the cases where it is genuinely warranted are narrow: a manual action reported in Search Console, or a documented paid-link history you are remediating. We have seen more damage done by over-eager disavow files than by the links they were written to neutralise. If in doubt, do not.

Where should the links point, the home page or inner pages?

Both, in a natural mix, but the more useful question is whether the page you want to rank has any equity pointing at it at all. A very common pattern on small sites is a respectable domain-level profile and a service page with nothing pointing at it, internally or externally. The fix there is internal linking, which you control entirely and which is usually faster than outreach. Fix that before you go looking for external links to a specific page.

My competitor has far more backlinks and I still outrank them. Why?

Because links are one input among many and the system evaluates relevance at query level. Their profile may be weighted towards sources that are heavily discounted, their pages may be a poorer match for the query, or their site may be harder to crawl. It is a useful piece of evidence rather than an anomaly: it tells you that in your niche, links are not currently the deciding factor, and your effort belongs somewhere with a better return.