How to get backlinks for a new website: we checked 200 launches

A new website gets backlinks it never asked for. A few months after launch, a typical one has dozens of referring domains, and most of them link to every new domain without anyone choosing to. So which links come on their own, which ones count, and how do you get the second kind?

We took 200 websites launched on Hacker News ("Show HN") between January and June 2026, on domains registered at most four months before the post. On October 31, 2026, we pulled their backlinks: 21,652 referring domains. This is what we found:

• Most backlinks of a new website are automatic. 46% of the referring domains are automatic SEO sites: "backlink checkers", "traffic checkers", URL shorteners. Their median authority is 1 out of 100. 186 of the 200 websites had at least one.
• A launch on Hacker News brings copies. 23% are sites that copy the Hacker News post or put it in a news digest. They are 69% of the referring domains first seen in the week after the post.
• What remains is small. A median of 19 referring domains per website are neither automatic SEO sites nor copies. A median of 5 of them have an authority of 30 or more, and 30 websites had none.
• Points on Hacker News go with more links. Websites whose post got fewer than 5 points have a median of 30 referring domains, 1 of them useful with an authority of 30 or more. With 100 points or more: 148 referring domains, and 13.
• Dofollow links are not what new websites lack. 83% of the referring domains give at least one dofollow link, and 93% of the automatic SEO sites do. What is missing is links from sites with readers.
• Directories you submit to are rare. At least one of the 199 directories in our catalog linked to 56 of the 200 websites. Launch lists and directories as a whole have a median authority of 8: many of them add Show HN posts on their own.

How we checked#

• The launches. Hacker News lists 25,959 Show HN posts between January 1 and June 30, 2026. Its search returns at most 1,000 posts per query, and we read 22,099 of them, week by week. 8,938 of those link to a website on its own domain. We left out GitHub, app stores and shared hosting addresses such as vercel.app, and kept the first post for each domain.
• New websites only. We read the registration date of 661 of these domains, taken at random, in the registries' public records. 376 were registered at most 120 days before their post. Their median age on the day of the post was 14 days. Domains whose registry publishes no registration date, such as .io, are not in the sample.
• Four groups. We took 50 websites at random in each group of points: fewer than 5, 5 to 19, 20 to 99, and 100 or more. Most Show HN posts stay in the first group: 80% of the 8,938 websites got fewer than 5 points.
• The backlinks. On October 31, 2026, we read each website's referring domains in SE Ranking's backlink index: the site, its authority (SE Ranking's score from 0 to 100, close to a Domain Rating), whether at least one of its links is dofollow, and the day SE Ranking first saw it. We read up to 250 referring domains per website. 42 websites had more: for them, we have the 250 most recent.

We sorted the referring domains into five kinds, by rules on the site's address and its home page title. We checked the 170 most frequent sites by hand.

• Automatic SEO sites: "backlink checker", "DA PA checker" and "traffic checker" sites, URL shorteners, and pages of keyword rankings or site statistics made for every domain.
• Copies of the Hacker News post: Hacker News clones and readers, sites that track Show HN posts, news digests and public RSS readers.
• Launch lists and directories: sites that list products, whether you submit them or they add them on their own.
• Communities and code: Hacker News itself, Reddit, forums, GitHub.
• Blogs, newsletters, media and other sites: articles, newsletters, project pages, company sites.

A backlink index does not see every link. SE Ranking sees few links on social networks and Reddit, so the share of communities is low. The day it first saw a link can come after the day the link was made. Sorting by rules makes mistakes: read the shares as orders of magnitude, not as exact figures.

How many backlinks does a new website get?#

Four to ten months after their Show HN post, the 200 websites had a median of 87 referring domains, from 4 to more than 250. The points of the post go with the number of links:

The points do not create the links on their own. A post that gets 100 points is on the front page, where more people read it, and some of them write about it. The posts that get points are also, often, the products people find interesting. Both count.

The total is the misleading figure. A website with 30 referring domains can have 1 that a person with readers chose to add.

Where the backlinks come from#

Automatic SEO sites: 46%#

These sites have names like trafficchecker.store, dofollowlink.space or backlinkchecker.website. They belong to networks of hundreds of domains on cheap extensions (.shop, .store, .space, .website, .online, .site). Each one publishes a page for every new domain it finds, often to sell backlink or traffic services. In the source pages we opened, the same page and the same link text appeared on dozens of these domains at once.

• Their median authority is 1. 93% of them give a dofollow link.
• 186 of the 200 websites had at least one, with a median of 6 per website.
• 90% of them were first seen in October 2026, in one wave that reached 134 of the 200 websites, whatever their launch month.
• Some arrive before the launch: 62% of the referring domains seen before the Show HN post were these sites. Some of them publish lists of newly registered domains.

You don't need to do anything about them. Google says that "in most cases, Google can assess which links to trust without additional guidance", and keeps the disavow tool (opens in a new tab) for sites with a manual action, or about to get one, for spammy links. Since December 2022, its SpamBrain system "can now detect both sites buying links, and sites used for the purpose of passing outgoing links", as Google wrote when it announced it (opens in a new tab).

They matter for one thing: they inflate the backlink counts in SEO tools. Count your referring domains without them.

Copies of the Hacker News post: 23%#

A Show HN post is copied within days:

• Hacker News clones and readers, many built as demos of a web framework, which show the same front page with the same links.
• Sites that track Show HN posts. One of them linked to 192 of the 200 websites, another to 180.
• News digests and public RSS readers in English, Chinese, Japanese, Korean and French, which rewrite or translate the post.

They are 69% of the referring domains first seen in the week after the post. Their median authority is 4, and 78% give a dofollow link. They show that the post was read, but each one repeats the same title and the same link. Nobody chose your product there.

Launch lists and directories: 11%#

Two kinds of sites are here:

• Launch lists that add products on their own. Several add Show HN posts without being asked: one listed 148 of the 200 websites, another 119. Most have an authority under 10.
• Directories you submit to. These have more authority and real visitors, and they don't come on their own. At least one of the 199 directories in our catalog linked to 56 of the 200 websites. The other 144 had none.

The median authority of this kind is 8, and 58% of its referring domains give a dofollow link. Among directories, the one that lists you without asking is rarely the one people read.

Blogs, newsletters, media and other sites: 18%#

This is where the authority is: 39% of these referring domains have an authority of 30 or more, against 7% of the automatic SEO sites. They are articles, project pages, company sites, and newsletters, the clearest source in our data:

• Newsletters sent with Buttondown linked to 31 of the 200 websites.
• Web Tools Weekly linked to 12, TLDR and Frontend Focus to 5 each.
• Hackaday, a media site, linked to 5.

A newsletter editor reads the launch, tries the product, and picks it for their readers. That is the link Google's own description fits: it "uses links as a signal when determining the relevancy of pages and to find new pages to crawl", as it says in its guide to links (opens in a new tab).

Communities and code: 2%#

SE Ranking recorded a link from Hacker News itself for 25 of the 200 websites, and from GitHub for 7. The share is small because a backlink index sees few links on social networks and Reddit.

When the backlinks arrive#

• Before the post, automatic SEO sites, which track newly registered domains.
• The first week, the copies of the Hacker News post.
• From the second week to the third month, articles, newsletters, other sites and directories: about half of the referring domains first seen in that period.
• After three months, the automatic SEO sites again, in the October wave.

A launch gives one short window. 61% of the blogs, newsletters and other sites that linked after the post did so within three months, while people still talked about the product.

Do nofollow links help SEO?#

Google treats nofollow as a hint, not as an order. Google wrote in 2019 (opens in a new tab): "All the link attributes — sponsored, ugc, and nofollow — are treated as hints about which links to consider or exclude within Search."

In our data, the attribute is not what separates links. The automatic SEO sites are 93% dofollow, more than any other kind, and their median authority is 1. A link from a newsletter or a directory people use, dofollow or not, is worth more than a dofollow link from a site that links to every new domain. Judge a link by the site it comes from.

How to get backlinks for a new website#

1. Launch where people read#

Show HN, Product Hunt and the other launch platforms bring the readers who then write about you. In our sample, websites whose post got 100 points or more had a median of 13 useful referring domains with an authority of 30 or more. Those under 5 points had 1. Hacker News's guidelines (opens in a new tab) say what helps: "Please make it easy for users to try your thing out, ideally without barriers such as signups or emails." They also say: "Please don't ask friends to upvote or comment."

A Show HN is one day. Our list of 72 launch platforms gives more chances to be read, each with its own audience.

2. Write to newsletters in your field#

Newsletters gave the clearest links in our data. Find the five to ten newsletters your users read: a newsletter for front-end developers, for designers, for marketers. Send the editor two or three lines: what the product does, for whom, and a link to try it without signing up. Many newsletters have a form or an address for suggestions.

3. Submit to the directories people use#

The launch lists that add you on their own are not the ones with authority. The directories with readers need a submission: a name, a description, images, sometimes a badge on your site. 144 of our 200 websites had no link from any of the 199 directories we track. Our article on directory submission lists them, with the link each one gives and how long each one takes.

4. Publish something others can cite#

A page with data no one else has, a free tool, an open-source project on GitHub, a clear comparison of the tools in your category. These are the pages other sites link to after the launch. Our study of what AI assistants cite found the same kind of pages: comparisons with real alternatives and prices.

5. Ignore the automatic links, and don't buy any#

Most of your first backlinks will be automatic SEO sites and copies. Leave them. Don't pay for "free backlink" packages or link networks either: Google's spam policies (opens in a new tab) list "Buying or selling links for ranking purposes" and "Low-quality directory or bookmark site links" as link spam.

6. Count the links that matter#

Once a month, open your backlink tool or Search Console's Links report and count the referring domains that are not automatic SEO sites or copies of your launch. In our sample, the median website had 19 of them after four to ten months, 5 with an authority of 30 or more. That number, and its trend, says more than the total.

The first steps of SEO for a new site, before and after the first links, are in our guide to SEO for a new website.

Where Submitium fits#

Submitium does one part of this: the directories you have to submit to. In our sample, 144 of the 200 websites had no link from any of the 199 directories in our catalog, and the lists that linked to them on their own had little authority.

We submit your product by hand to directories and launch platforms that give a free dofollow link, keep their badges in one badge wall on your site, and check the link on each live listing six times over 90 days. A directory costs $0.99, less from the 51st, and a first order starts at 30 directories. See the pricing and the full list of directories.

Questions#

How do I get backlinks for a new website?#

Launch where people read (Show HN, Product Hunt and other launch platforms), write to the newsletters of your field, submit to the directories people use, and publish something others can cite. In our study of 200 new websites, newsletters, articles and directories gave the links with authority. Most other links came on their own and were automatic.

How many backlinks does a new website need?#

There is no fixed number. In our study, four to ten months after launch, the median website had 87 referring domains, of which 19 were not automatic and 5 had an authority of 30 or more. A few links from sites with readers count more than a large total.

Are free backlinks worth it?#

The sites that offer "free backlinks" are mostly the automatic SEO sites of our study: median authority 1, linked to almost every new domain. Google lists low-quality directory and bookmark links as link spam. Free links from sites people read, such as a newsletter or a directory with visitors, are worth having.

Should I disavow spammy backlinks?#

In most cases, no. Google says it "can assess which links to trust without additional guidance" and recommends the disavow tool when spammy links caused a manual action, or likely will. 186 of our 200 websites had automatic SEO sites linking to them.

Do nofollow links help SEO?#

Google treats nofollow as a hint about which links to count, not as a rule. A nofollow link from a site with readers also brings visitors. In our study, the automatic SEO sites were the most dofollow kind, 93%, with a median authority of 1.

What are referring domains?#

The different websites that link to yours. Ten links from one site are one referring domain. SEO tools count them because one link from each of 20 sites says more than 100 links from one site.

How long does it take to get backlinks after a launch?#

Copies of a Show HN post appear within days: they were 69% of the referring domains first seen in the first week. Articles, newsletters and directories came mostly between the second week and the third month. Search Console's Links report can take weeks to show them.

What is SaaS link building?#

Getting links to the site of a software product: from launch platforms, directories, newsletters, comparison pages, integration partners and articles. For a new SaaS, the first links come from the launch and from directories; the later ones from pages that compare tools in its category.