How do I get backlinks when nobody knows my site?

Every guide to link building assumes a starting point you do not have. Broken link building, resource page outreach, digital PR — all of it presumes that when someone visits your site to evaluate the request, they find something that makes linking seem reasonable.

At zero authority, with fifteen pages and no traffic, they do not. Which is why new sites run outreach campaigns with response rates near zero and conclude that link building is a scam, when the actual problem is running step three before step one.

The order most people get wrong

Link building is the most expensive form of SEO, in money and in time, and it is the one most commonly attempted first.

The reasoning is understandable — everyone knows links matter, and it feels like the thing that separates you from the sites above you. But it inverts the economics. On-page and technical work costs little, is entirely within your control, and reports back within weeks. Link acquisition costs a great deal, depends on other people, and reports back in months.

So the sequence that works is: fix indexation, choose winnable queries, build genuinely good pages, link them internally, then establish whether you have an authority problem at all. Most sites never reach that last step, because the cheaper work solved the problem. That progression is laid out in our step-by-step SEO guide.

You have an authority problem when — and only when — you can name specific queries where you rank between five and fifteen with a page that is genuinely better than the ones above you. Anything else is a relevance problem wearing an authority costume, and links will not fix it.

Six approaches that work from zero

1. Publish data nobody else has. The single most reliable method, and most companies are sitting on the raw material without realising it. Whatever your product does, it generates observations about your industry that nobody outside can make. Aggregate it, anonymise it, publish it with a clear methodology. Journalists and bloggers need statistics to cite, and there are far fewer sources of original numbers than there are people needing them.

This works at zero authority because the link is not a favour. Someone writing an article needs a figure, yours is the only one that exists, and they cite it. Authority is irrelevant to that transaction.

2. Build one small free tool. A calculator, converter, checker or generator that solves one specific annoyance in your field. Tools accumulate links continuously because people reference them in answers, forum replies and documentation for years. The bar is lower than it sounds — most successful link-earning tools are a single page doing one small thing well.

3. Take a clear position on a live debate. Every industry has questions practitioners argue about. Most published content dodges them to avoid alienating anyone, which is precisely why a well-argued position gets referenced — as agreement or as the thing being disagreed with, both of which link.

4. Be genuinely useful in communities. Not link-dropping, which is transparent and usually removed. Answering questions properly in the places your audience already gathers, with your site as context rather than pitch. These links are often nofollow and still valuable, because the pages they sit on are retrieved heavily by AI assistants — a mechanism covered in which platforms get cited in ChatGPT.

5. Get listed where your category is catalogued. Industry directories, comparison sites, curated lists, association memberships, integration and partner pages of the tools you work with. Unglamorous, finite, and mostly achievable in a week. Partner pages in particular are routinely overlooked — every vendor you integrate with likely has a page listing companies like yours.

6. Reclaim mentions you already have. Search for your brand name and look for places you are named without a link. Someone has already decided you are worth mentioning; asking them to link the mention is the highest conversion rate in outreach by a wide margin.

What actually earns a link

The common factor across everything above: the page gives someone a reason to reference it that has nothing to do with helping you.

People link for three reasons. To support a claim they are making, which needs data or a definitive statement. To save their reader effort, which needs a tool or a genuinely comprehensive resource. Or to argue with something, which needs a position.

Notice what is absent. Nobody links because a page is well written, thorough, or nicely designed. “Great content earns links” is repeated constantly and is not true as stated — excellent content earns links when it happens to contain one of those three things, and enormous quantities of excellent content earn nothing because it contains none of them.

So before any outreach, ask: what is on this page that someone could not get elsewhere, and what claim does it let them make? If there is no answer, the outreach will fail and the failure will be correctly attributed to the page rather than the email.

Outreach that gets replies

Once you have the asset, outreach becomes a matter of finding people who have already demonstrated they want it.

The highest-yield targets, in order:

• People who linked to a worse version. Find who cites the outdated study or the abandoned tool your work supersedes. They have already made the editorial decision; you are offering an upgrade.
• People who mentioned you unlinked. Highest conversion rate available.
• People writing about the exact subject now. Recency matters enormously — an article published this week is still being edited; one from 2023 is not.
• People whose article has an obvious gap your data fills, where you can point at the specific paragraph.

The message itself matters less than the targeting, but three things reliably improve replies: be specific about which page and which paragraph, state the value in one sentence without adjectives, and do not ask for anything beyond consideration. Templates that open with praise for their “insightful article” are filtered before they are read.

Expect single-digit percentage responses even when everything is right. Fifty well-targeted emails producing four links is a good campaign, and four relevant links is meaningful movement for a new site.

Three that waste the budget

Paid links and private blog networks. The policy risk is real but secondary. The practical problem is that sites selling links have no genuine audience, and a link from a site nobody reads carries little weight regardless of its metrics. You are buying a number in a tool rather than an outcome.

Mass guest posting. Not the genuine kind — writing something substantive for a publication with real readers remains worthwhile. The failure mode is the industry of sites that exist to host guest posts, which have no readership and whose pattern is easy to identify. The test: would you still want this placement if the link were nofollow? If not, it is not worth the fee.

Directory submission at volume. A handful of genuine industry directories, yes. Hundreds of general-purpose ones, no. This was effective in 2009 and has not been for a long time, but it persists because it is easy to sell as a package with an impressive-sounding quantity.

A useful heuristic across all three: if the method scales cheaply, it does not work. Anything that can be bought in bulk has been bought in bulk, and detection followed. Methods that resist scaling — original data, real tools, genuine relationships — keep working precisely because they cannot be commoditised.

How to tell whether it is working

Not by counting links. Count referring domains, which is the number that matters — fifty links from one site is roughly one site’s worth of signal.

Then watch the metric that actually indicates progress: whether pages that were stuck outside the top ten begin moving. Authority is not an end in itself; it is the input that unlocks positions relevance alone could not reach. If referring domains are rising and stuck pages are not moving, the links are probably irrelevant to those queries — a common outcome when link acquisition is run as a volume exercise disconnected from the pages that need help.

Compare against the sites you are trying to displace rather than against an absolute target. A competitor analysis tells you the gap in referring domains for the specific queries you care about, which converts an unbounded task into a finite one.

Two further habits worth adopting early. Record where every link came from and what prompted it, because the pattern tells you which of the six approaches is working for your specific field — it is rarely the same one across industries, and after twenty links the answer is usually obvious. And check quarterly for links you have lost: pages get deleted and articles get rewritten, and a link reclaimed by a short email is cheaper than a link earned from scratch.

One final note on where this is heading. Assistants selecting sources for an answer lean on being referenced across the places they retrieve from, but they weight passage relevance and freshness far more heavily than accumulated authority. For a new site this is genuinely good news: it is a channel where being precise and current can outcompete being established, and it is available months before the link math works in your favour.

Frequently asked questions