How to Get Backlinks
Backlinks are links from other websites to yours, and they remain one of the strongest signals Google uses to decide which page deserves to rank. You earn them by making something worth referencing, then telling the right people it exists.
There is no safe shortcut. Buying links, exchanging them in schemes, or mass-posting them violates Google's spam policies and risks the site you are trying to help.
This is the least satisfying area of SEO, because the honest advice is slow and cannot be automated. That is precisely why it still works — everything easy has been done to death and neutralised.
Why they matter
A link is a vote that costs the linker something. They are staking a little of their own credibility on you being worth a reader's click, which is why links have survived twenty-five years of attempts to game them.
For a new site this is usually the real ceiling. You can write better than a competitor and still lose to them, because a domain nobody references has to prove itself from nothing. After the on-page work is done — good titles, internal links, genuinely useful content — links are what is left.
What actually works for a small site
Be genuinely useful somewhere people ask questions
Reddit, Stack Exchange, niche forums, Discord servers, Facebook groups. Answer the question fully in the reply itself, then link only if a link genuinely adds something.
Most of these links are nofollow and pass little direct ranking signal. They are still worth it: they bring real readers, and readers with their own sites are where genuine links come from. A drive-by link drop gets deleted and earns you nothing.
Publish something nobody else has
The most reliable way to earn links without asking. Original data is the strongest form — a measurement, a survey, a test you ran, numbers from your own work.
This site's traffic article publishes its own Search Console figures: 90,700 impressions producing 26 clicks, and top-10 conversion queries earning 0.17% click-through. Those numbers cannot be found anywhere else, because nobody else has them. Whether people cite it or not, the reason to try is that only unreplicable material gives someone a reason to link.
You do not need a research budget. You need one thing you measured that everyone else guesses at.
Respond to journalists and writers
Reporters and bloggers actively look for people to quote. Services exist that circulate those requests to sources. Answering promptly, specifically, and within your genuine expertise gets you cited with a link — from exactly the kind of site whose links carry weight.
The success rate is low and the effort per pitch is small. It suits people who genuinely know a subject and can write two useful paragraphs quickly.
Guest posts, done properly
Still legitimate when the article is genuinely good, on a site with real readers, on a topic you know. What Google objects to is guest posting at scale purely for links — spun articles placed anywhere that accepts them.
The test is simple: would you write this piece if the link were removed? If not, do not write it.
Fix the links that already exist
Two easy wins people miss. Search your brand name — you may already be mentioned without a link, and a polite email often converts those. And check for broken links pointing at your site after a URL change; a redirect recovers the value immediately.
What to avoid
| Tactic | Why not |
|---|---|
| Buying links | Direct spam-policy violation. Sellers reuse the same networks Google already knows. |
| Link exchanges ("I'll link you if you link me") | Explicitly named in the policies. Trivial for Google to detect. |
| Private blog networks | Sites built solely to link. Deindexed in waves, taking clients with them. |
| Mass directory submissions | Worthless since the 2000s. A few reputable niche directories are fine. |
| Comment and forum spam | Nofollowed, deleted, and damages your name with the exact audience you want. |
| Automated outreach at volume | Ignored. Recipients recognise the template instantly. |
The economics of buying links do not work. Anyone selling a link at scale is selling it to everyone, which is how those networks get identified. You inherit the risk; they keep the money.
Asking, without being ignored
Most outreach fails because it is obviously mass-sent. What helps:
- Give a reason that is about them. Which page, why your resource improves it for their readers.
- Prove you read the page. One specific, accurate sentence does more than any amount of flattery.
- Be short. Three or four sentences.
- Ask once. A single follow-up after a week or two is acceptable; more is not.
- Accept no. Silence is an answer.
Expect a low response rate even when you do this well. Ten thoughtful emails beat two hundred templated ones, and cost less time overall.
Quality over quantity, concretely
One link from a site your readers already trust is worth more than a hundred from sites nobody visits. What makes a link valuable:
- Relevance — a link from a site in your subject area means more than an unrelated one.
- The linking site's own standing — do others reference it?
- Editorial placement — inside the body of an article, not a footer or sidebar list.
- Descriptive anchor text, occurring naturally. Repeating an exact keyword across many links looks manufactured, because it is.
A realistic timeline
In your first year you may earn a handful of links, and several will come from places you did not plan. That is normal. Link building compounds: the first is hardest, and each one makes the next slightly easier as more people encounter your work.
If you have no links after six months of publishing, the usual reason is not that outreach failed — it is that nothing published so far gives anyone a reason to link. That is a content problem wearing a link-building costume.
Frequently asked questions
How do I get backlinks for a new website?
Publish something worth referencing — original data is best — then participate genuinely where your audience asks questions, and reach out individually to sites whose readers would benefit. Expect a handful in the first year, not hundreds.
Are backlinks still important in 2026?
Yes. They remain one of the strongest ranking signals, particularly for a new domain with no track record. Their relative weight has fallen as content quality signals improved, but they have not been replaced.
Is buying backlinks safe?
No. It directly violates Google's spam policies, and sellers reuse the same networks across many customers, which is how those networks get identified and neutralised. You carry the risk; the seller keeps the fee.
How many backlinks do I need to rank?
There is no number. A handful of relevant links from trusted sites can outperform hundreds of low-quality ones. Look at what currently ranks for your target query rather than chasing a total.
Do nofollow links help?
Not much directly — Google treats them as a hint rather than an endorsement. They are still worth having, because they bring real readers, and readers who run their own sites are where followed links originate.
Is guest posting still allowed?
Yes, when the article is genuinely good and on a site with real readers. What breaches the policies is guest posting at scale purely to place links. If you would not write the piece without the link, do not write it.
How long does link building take to show results?
Months. A new link takes time to be crawled, and its effect on rankings is gradual rather than immediate. Treat it as compounding over a year, not as a lever you pull for next week.
Conclusion
Links are earned by being worth referencing and then being findable. Original measurements, genuine participation where your audience already is, and a small number of well-judged asks will out-perform any volume tactic — and none of it puts the site at risk.
If nobody is linking after months of effort, the honest question is not how to get more links. It is what you have published that anyone would want to point at.
Related: getting traffic to a new blog covers the wider picture, finding blog post ideas covers choosing topics worth writing, and why a site is not showing on Google covers the indexing side.
Comments