How to get backlinks in 2026 (that actually count)

Ask ten SEO people how to get backlinks and you’ll hear ten versions of the same list: guest posts, directories, broken links, a few outreach templates, and a vendor pitch. Some of those still work. A few now cost more than they return. The useful question isn’t which tactics exist. It’s which ones still produce links that pass real value, and what each one costs you in time and exposure. This playbook ranks the methods from the ones that compound down to the ones you should skip.

Here’s the counter-intuitive part. The best link methods don’t feel like link building. They look like making something a stranger would cite, then telling the right person it exists. The worst ones feel like link building because the link is the entire point of the page.

What a backlink has to do before it counts

A backlink only helps when four things are true. Search engines have to discover the linking page, crawl it, and index it. The link has to be a real, followable <a href> inside the page’s content. The source has to be connected to your topic. And the placement has to look like an editor chose it.

The first condition trips up more people than any other. A link on a page that never gets indexed does nothing for ranking, however good the site looks from the outside. Google has treated rel="nofollow" as a hint rather than a directive since 2019, and it recognizes rel="sponsored" and rel="ugc" for paid and user-generated links. Those attributes tell Google how to treat a link. They don’t decide whether the link helps you. A nofollowed link from a busy, relevant site can still send real readers and brand exposure, and a dofollow link from a dead page can do nothing at all.

Indexing is the step most link reports ignore. Think of the whole process as a four-step loop: find a page that fits your topic, earn the placement, confirm the link is live and followed, and then confirm the linking page is indexed. Step four is the one people skip, and it’s the only one that tells you whether the link is doing anything.

Relevance is the other half of the mechanism, and we made that case in relevance beats volume. A link from a page about your topic carries context. Context is what makes a link believable to readers and, as far as anyone outside the search companies can tell, to the systems that weigh it.

How to get backlinks, ranked by effort, risk and durability

Each method below gets scored on four things: how relevant the resulting link is, how much effort it takes, how much risk it carries, and how long the link lasts. Durability matters more than most people account for. A link that stays for five years is worth many that vanish after six months.

The top of the table is where the work compounds. The bottom is where it gets you penalized. The sections below run the top methods in order and explain why the bottom ones fail.

Start with the links and relationships you already have

The cheapest links are the ones someone already decided to give you. Writers mention brands all the time without linking to them. A reviewer names your tool in a roundup, a podcast host puts you in the show notes, a newsletter describes your category and drops your name. Each one is a warm mention with one link missing.

Find them with a search for your brand name, excluding your own domain: "yourbrand" -site:yourdomain.com. Run variants for your product name, your founder’s name if it’s public, and common misspellings. Run it monthly or set up alerts. Brand-monitoring tools do this continuously, but their coverage is patchy, so the search operators are your baseline.

Then write the request. Keep it to three sentences: thank them for the mention, name the page where the link would go, and ask. A one-line request is easy to say yes to, and most people who wrote about you are glad you noticed. Send one follow-up if they don’t answer, then stop. Chasing someone who ignored two emails spends goodwill you’ll need for the next one.

Your own history holds a second source. Pages that still link to a URL you’ve since changed are links that quietly broke. Check your most-linked pages for 404s in Search Console or a crawler. A 301 redirect keeps visitors arriving and tells search engines where the page moved, but a direct link to the new URL is cleaner, so ask the linking sites that matter to update it.

Partners and customers come third, and they’re the most underused category for SaaS companies. Every integration you ship is a relationship with another company that already has a page about it and an audience searching for it. Their integration directories and partner pages are relevant to you by definition, because they describe your product.

Customers are just as valuable. A customer who publishes a case study on their own blog, with results from their own operation, gives you a link on a page about the exact problem you solve, and the anchor is naturally your name. The ask isn’t “please link to us.” It’s closer to “would you like to write up how you cut onboarding time?” Then do most of the drafting yourself, so the customer only has to approve it. Give something back, such as a co-published article, a feature in your newsletter, or a joint webinar. Customers who want their own visibility will say yes. Customers who only wanted a discount will not, and you’ll learn that before you’ve written anything.

Build assets other sites need to cite

Editors link to things they couldn’t get anywhere else. That’s the logic behind the strongest link asset most companies can build, which is original data or a tool that solves a narrow problem. Both give a writer something to cite, and citation is the whole game.

Original data

A journalist writing about remote hiring needs a number. If you have a number nobody else has, with a method you can describe, you become the citation.

Take a project-management tool with a few thousand customers. It can publish a report on how much time teams spend in status meetings, drawn from anonymized calendar data of accounts that opted in, sliced by team size. Competitors can’t copy that dataset, and the finding is written in a form writers can quote. The report needs a methodology section, a sample size, and a date range. Without those, editors skip it, and a few will publicly question it.

Surveys work the same way and are easier to start. A few hundred respondents in one specific job title is enough to publish, provided you disclose how you recruited them and what you asked. A broad survey of “marketers” is a weaker asset than a survey of agency owners about one workflow, because the narrow version answers a question a specific writer is already asking.

Pitch the data to the writers who cover that beat, with one finding in the subject line and a link to the page. Don’t attach the full report. Most pitches will get nothing. A small number will produce links that keep paying for years.

Free tools

A calculator, generator or checker becomes a link magnet when it solves one small, recurring problem in your niche. Invoicing software can publish a late-fee calculator, which is narrow and useful. A generic “ultimate SEO toolkit” is neither.

The links come from roundup and resource pages, which list tools in a category. Find those pages first. Search for best [category] tools and free [category] tools, see which pages already rank, and then ask their authors to add yours. The pitch is short: here is a free tool that does one specific job, and here is how it differs from what you already list.

Two costs people underestimate. The tool has to keep working, which means maintenance indefinitely. And it has to be correct. A calculator with a wrong formula earns links for a month, then gets corrected in public by the same people who linked it.

Outreach that earns placements

Outreach works when you give someone a reason to change their page. Three methods do that reliably, and each one needs a different kind of work.

Broken link building

Broken link building works because it solves a problem for the site owner. A resource page links to something that now returns a 404, and you have a replacement. The editor fixes their page without doing research, and you get a link on a page that’s already about your topic.

Run it in this order:

1. Find resource pages in your niche that list references, tools or guides. Crawl them with a tool like Screaming Frog, or use the broken-link report in an SEO platform, to find outbound links that return errors.
2. Check what the dead link used to point to. The Wayback Machine shows the original page, which tells you what the editor was linking to and what a replacement needs to match.
3. Decide whether you have a page that genuinely matches. If you don’t, don’t pitch. A weak substitute reads as a link request dressed up as help, and editors can tell.
4. Email the editor with the exact broken URL, the page where it appears, and your replacement. Keep the note to four or five sentences and name the specific link.
5. Log every result, including the ones that never reply. Hit rates are low, so keep the list tight and relevant. One broken link on a page that matches your topic is worth more than a hundred on unrelated sites.

Journalist expert requests

HARO is gone, and a set of replacement services now routes reporter queries to sources. The structure is the same as the old model: a reporter needs an expert on a specific question, and you answer with something specific and sourced. A link appears when the reporter cites you and the outlet links to the source.

This is cheap per pitch and less durable than a data study, and some links come back nofollowed. Treat it as a steady channel, not a link program. The discipline that matters is honesty. Answer only the queries where you actually know the answer, and answer quickly, because many close within hours. A reporter who gets a vague quote from you won’t come back, and reporters talk to each other.

Editorial guest posts, the narrow version

Guest posting is the method most damaged by bad practice, and it still works when you treat it as writing for a publication rather than buying a link. The test is blunt: would you want this post on that site if the link didn’t exist? If not, skip it.

A good placement has a clear editorial focus that overlaps with your buyers, an editor who actually edits, and a post about one specific problem. Your link belongs in the body, where it supports a claim. An author bio can carry one link, and that’s enough.

The red flags are easy to spot. A site that accepts posts on forty unrelated topics. A “write for us” page that lists a price per link or a minimum domain rating. A pitch that offers the same article to every publication in a category. If a publication will sell you a link, it is selling links to everybody.

What’s dead, and why it keeps getting sold

• Bulk directory submission. Generic directories with hundreds of categories and no editing create the footprint of a manipulation campaign. A curated niche directory that people in your field actually use is a different thing.
• Profile and article spam. Profiles on free blogging and Q&A platforms, stuffed with links and rewritten articles, produce the pattern search engines are built to recognize, and they rarely hold value for long.
• Reciprocal link exchanges at scale. “You link to me, I link to you” arrangements create pages that exist only to point outward. Two partners linking each other from a real partner page is fine. A page that exists to hold forty swaps is not.
• Sitewide and footer placements sold in bulk. A link that appears on every page, sold by the month, looks engineered because it is.
• Private blog networks. Networks of sites built to pass links are the clearest form of link scheme. Google’s spam policies name link schemes directly, and when a network gets identified, its links tend to disappear together, taking any ranking they supported with them.
• “Guaranteed” link packages. A promised number of high-authority links for a flat fee is the giveaway. Nobody can guarantee editorial placements on sites they don’t control, and what gets delivered at that price is usually the bottom of the market.

These still get sold because they’re easy to buy and easy to report. A dashboard full of new links looks like progress for three months. That’s the reason to be skeptical of any vendor whose headline number is link count.

How to get backlinks without getting burned

Most link penalties aren’t about one link. They’re about a profile that looks built rather than earned. Four things tell the story.

The shape of your growth

Natural link growth is lumpy and continuous. A piece gets picked up, nothing happens for weeks, then a second wave arrives when someone cites it. Manufactured growth is a vertical spike from nothing, followed by a flat line. No published number tells you a safe count of links per month, and anyone who gives you one is guessing. Judge the shape, and judge the sources behind it.

Anchor text

Anchors should match how people naturally refer to you: your brand, your domain, the topic of the page, and generic phrases like “this guide.” If a large share of your links carry the same keyword-heavy anchor, that’s the pattern to fix. The test is whether a real writer would phrase the sentence that way. “Best link building software” dropped into a random blog post about contractors fails it.

The source

Before accepting a link, look at the site. Does it get organic search traffic on its own topic? Does it have named authors and a stated editorial policy? Does it link out to sources that make sense? Third-party traffic estimates are rough, so use them to rule out sites that clearly have none, not to rank the close calls.

Paid and sponsored placements

If money or free product changed hands, the link should carry rel="sponsored" and the page should disclose the arrangement. Google’s guidance on paid links asks for exactly that, and readers are entitled to know. Unlabeled paid links are the kind that get manually actioned.

The disavow file

Google’s disavow tool has one job: asking Google to ignore links you can’t get removed, usually when you have a manual action for unnatural links or are confident a pattern of spam is causing harm. It isn’t routine hygiene. Teams that disavow links that were fine remove real value and gain nothing. Audit first, separate the clearly manipulative links from the merely unfamiliar ones, and disavow only the first group.

A 90-day sequence for a site with few links

1. Weeks 1 and 2: claim what’s already yours. Run the brand-mention search, send the requests, and fix the 404s on your own pages that other sites link to. This takes an afternoon and often produces the first links you’ll get all quarter.
2. Weeks 2 to 6: build one linkable asset. Choose one thing: a dataset from your product, a survey of a specific job title, or a free tool for a narrow problem. Ship it before any outreach. Outreach without something worth linking to is just email.
3. Weeks 4 to 8: work through partners and customers. List every integration partner and every customer with a result they’d talk about. Pitch a small number from each group, and bring a draft they only need to approve.
4. Weeks 6 to 12: run small, checked outreach. Start broken-link and resource-page outreach in batches you can verify by hand. Look at every target page yourself before you pitch it.
5. Ongoing: keep one channel for expertise. Answer journalist requests where you have a real answer, and publish one guest post a quarter on a site you’d read anyway.

What to take from this

None of this is fast. The methods that compound take a quarter before they show up, and a data study can land a handful of links and keep working for years. The methods that look fast mostly stop working once the pattern is noticed. The reliable answer to how to get backlinks is slower and less glamorous: make things worth citing, ask the people who already like you, and check that every link is indexed and still there.

That’s also why OptiLink is built around relevance rather than raw volume. Finding placements that fit the content, placing them where the link belongs, and then confirming each one is indexed and followed is slow to do by hand across a whole site. OptiLink finds relevant placements from millions of blogs, hand-places them where the link fits, and checks each one for dofollow and indexing, with live tracking so you can see what’s working. The easiest way to judge it is to see it run on your own site. Start with how it works, or browse the features.

Keep reading

Playbook

Anchor text optimization without tripping a filter

A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.

Sofia Meyer September 24, 20262 min read

Playbook

Broken link building, step by step: find, fix, and pitch

A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.

Sofia Meyer October 9, 20262 min read

Playbook

Competitor backlink analysis that finds real opportunities

A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.

Sofia Meyer September 26, 20262 min read

Explainer

Dofollow vs nofollow in 2026: what still passes value

The link attributes changed, Google's treatment changed, and a lot of advice is out of date. Here's what dofollow, nofollow, sponsored and ugc actually do now.

The OptiLink Team October 10, 20262 min read