three small source pages with arrows into one larger violet-marked page, a green check on the first arrow and a dashed fourth page still to be contacted

how to get backlinks: 6 methods and what they got us

a ranked process for earning links without buying them, with our own results: 10 list submissions, 4 merged, 4 closed, and what those links turned out to be.

crawlgraph team
· 11 min read · 2,330 words
sharexlinkedin

to get backlinks, find sites that already link to your competitors, give them a specific reason to link to you, and publish something worth citing. do those in order, starting with a list of real sites instead of a blank page. buying links is the shortcut to avoid. we also rechecked our own attempts for this guide, so you will see what each method actually returned, including the rejections.

if you are not sure what link building is, start with the definition. if you want to know how many links you are aiming for, we measured 580 first-page results.

  1. Choose the page and the target. Pick one page that deserves links and count the referring domains of comparable pages ranking for its keyword.
  2. Run a link gap. List the sites linking to two or three competitors and not to you, then keep the ones a person would actually visit.
  3. Get listed where your category is already compared. Submit to the few directories, resource pages and open-source lists that your buyers or peers actually use.
  4. Reclaim what is already yours. Ask for links on pages that mention you without linking, and for fixes on links that point at your dead urls.
  5. Publish something people need to cite. Give others a source: original data, a free tool, or a reference page nobody else maintains.
  6. Write to one person at a time and keep a ledger. Send a short message with a specific reason, follow up once, and record every send and result.

what these methods returned for us

crawlgraph is a small product with a small link profile, so our sample is modest. dataforseo's index listed 90 referring domains for crawlgraph.com when we read it on 30 october 2026. here is what we hold on record, and how recently we checked it.

methodwhat we didlatest result
open-source lists10 pull requests, rechecked on 30 october4 merged, 4 closed without merging, 2 still open
directories4 live listings in our trackeron 30 october, saashub had a followable main link and tiny startups was nofollow; the other 2 were not rechecked
developer articlesposts on dev.to with a link backdev.to was among our referring domains on 30 october
community commentshelpful answers on reddit and dev.toone removed by a karma filter when logged in may; the rest not rechecked

two things surprised us. first, the rejection rate: list submissions closed exactly as often as they merged. second, the link attribute. the github readme link we checked on awesome-osint carries rel="nofollow", which is a hint to google rather than an endorsement. the list's value is the people reading it and the sites that copy from it: trackawesomelist and awesomelist.link both appear in our referring domains. we also picked up scraper pages we never asked for, which is normal and not something to count as success.

what we cannot tell youwe did not measure traffic or sales per link, so nothing here shows that a given method paid for itself. treat this as a record of which links we could get, not what they were worth.

1. start with a link gap

a link gap is the list of sites linking to your competitors and not to you. it is the best place to start because the list is finite, and every site on it has already linked to something in your category.

the free crawlgraph gap analysis compares your domain against competitors using a named common crawl snapshot. it needs an email, returns the first 5 gaps and is limited per ip; a paid plan returns the full list. it works at domain level, so it tells you which sites to look at, not which page holds the link or whether it is still live. you can do the same job by hand, and free research methods covers how. then use this worksheet to record what you find.

sort by how many of your competitors each site links to. then open the top candidates and throw out anything a real person would never read. a site linking to all three competitors usually has a reason you can match, such as a comparison page or a resource list.

2. get listed where your category is compared

be selective here, because this method is easy to overdo. google's spam policies name “low-quality directory or bookmark site links” as an example of link spam. a good test: would anyone use this page to choose a product? software review sites, curated industry lists and active open-source lists can pass. a site that exists to sell listings will not.

for open-source lists, read the contribution rules and match the existing format exactly. resource pages work the same way: find them with searches such as "your topic" inurl:resources, then suggest your page only when it adds something the list is missing.

3. reclaim what you already earned

these are the easiest yeses, because the site has already chosen to talk about you. there are two common cases.

  • unlinked mentions. search for your brand or product in quotes, exclude your own site, and look for pages that name you without linking. a short note with the exact sentence and url is usually all you need.
  • broken or moved links. if a page links to a url of yours that now fails, redirect it or ask for the fix. the link reclamation guide covers the process.

broken link building is the same idea aimed at someone else's dead page: find a resource that has disappeared, then offer your replacement to the sites still linking to it. it only works if your page really does replace what was lost.

4. publish something people need to cite

most writers link when they need a source for a claim. so give them one: original data, a free tool, a calculator, or a reference page that nobody else keeps updated. a “complete guide” rarely does this, because there are already dozens.

look for a number people keep repeating without a source. our backlink count study came from that: search results kept repeating a 50-to-100 rule, so we measured it, and only 7.5% of pages on hard keywords fit. our who links to every seo tool study takes the same approach using the graph. it is too early for us to claim these have earned links, and a study takes more work than any other method here. it is also the only one that gives people a reason to find you.

5. write to one person at a time

outreach fails when the message could have gone to anyone. before you write, you should be able to name the exact page, the exact place your link would fit and why a reader of that page would benefit. if you cannot, skip the site.

txt
subject: [their page title] and a source for [specific claim]

hi [name],

[their page] says [specific sentence or section]. i noticed
[the exact gap: a dated figure, a missing option, a dead link].

we measured this recently: [one finding, one number].
[url of the page]

if it is useful, it would fit in [their section]. if not, no
need to reply. i won't follow up more than once.

[your name], [role] at [company]

keep it small and honest. send a handful of messages, not a blast. say who you are, make it easy to ignore, and follow up once at most. in many countries, unsolicited business email also has legal requirements, so check the rules that apply to you and honour every request to stop.

keep a ledger with the site, page, person, date sent, reply and final link status. that is what lets you check later that a link is still there, and stops you writing to the same person twice.

6. use guest posts and expert quotes for reach

guest posts can be worthwhile when the audience is one you want. the link is secondary. google names “links with optimized anchor text in articles, guest posts, or press releases distributed on other sites” as link spam, so avoid keyword-rich anchors and templated posts sent to many sites. write something you would be happy to publish on your own blog, and pitch the sites your buyers read.

answering journalist requests through a sourcing service can also earn a citation. we have no results of our own to report here, so we are describing the idea without vouching for it.

what to avoid

the same google policy lists the patterns worth skipping, including:

  • buying or selling links for ranking purposes, including trading goods or services for them. a paid link should be marked rel="sponsored".
  • excessive link exchanges, and partner pages that exist only to cross-link.
  • automated programs or services that create links to your site.
  • forum comments with optimized links in the post or signature.

if you find poor links you did not build, you usually do not need to panic. our guide to toxic backlinks and disavow explains when action makes sense, and how to judge backlink quality helps you decide which links are worth pursuing in the first place.

frequently asked questions

faq

what is the best way to get backlinks?

start with a link gap: find sites that already link to several of your competitors and not to you, then give each one a specific reason to add you. those sites have shown they link to your category. for a page worth citing, publish original data or a useful tool, because people need a source to link to.

how can i get backlinks for free?

the methods that cost only time are relevant directories and resource pages, open-source awesome lists, reclaiming unlinked mentions and broken links, and publishing something others want to cite. the cost is your time. google treats paying for links that pass ranking credit as link spam, so a free link that is earned is also the safer kind.

how do i get backlinks for a new website?

begin with the places a new product can legitimately appear: a few relevant directories, founder or company profiles, and lists that accept submissions. then run a gap against two or three competitors and write to the best matches. expect a slow start, and aim your content at low-difficulty keywords while the link profile grows.

how much do backlinks cost?

earning them costs time rather than a fee per link. prices for bought placements vary too much to quote usefully, and links bought for ranking purposes break google's spam policies unless they are marked rel=sponsored or rel=nofollow, which tells google not to treat them as an endorsement. budget for content and outreach time instead.

are backlinks still relevant in 2026?

yes. google still describes links as a way it finds pages and a signal of relevance, and links help people and crawlers discover your site. volume alone is a weak target, though: in our sample of first-page results, the median page had 5 referring domains, ranging from 2 to 48 by keyword difficulty.

monthly backlink toolslocked
subscription required
example backlinks $99 once
G
github.io
C
css-tricks.com
L
lobste.rs
A
algolia.com
common crawl graph · one-time
$99oncelifetime access
unlock the data →
stripe checkout · instant access
guides#link building#backlinks#outreach
sharexlinkedin
crawlgraph team
author

documents crawlgraph backlink workflows, open-data methods, and the limits of each report.

the dispatch
one email a month.

plus one when a new common crawl release lands. that is all.