Learning link building, from the earning side

Which links does Google call spam?

Google defines link spam as creating links to or from a site primarily to manipulate rankings. Its list of examples is blunt and worth reading in full: buying or selling links for ranking purposes, excessive link exchanges, using automated programs to build links, requiring a link in a contract, advertorials where payment buys a link that passes ranking credit, low-quality directory links, keyword-rich links buried in widgets, and links spread across the footers of many sites.

Right after that list sits a sentence almost nobody quotes. Google writes that buying and selling links is a normal part of the economy of the web for advertising and sponsorship, and is not a policy violation, as long as the link is qualified with rel="nofollow" or rel="sponsored". The red line is not money. It is passing ranking credit through a paid link.

So the question clients ask, "is buying links risky", was badly posed from the start. The real question is whether this link is meant to pass credit, and who decided that on purpose. If the seller marks it correctly you bought an advertisement and not a ranking; if they do not, the risk sits with your site rather than theirs. How the Persian market packages that rule is a separate argument, written up in our piece on buying backlinks, and not repeated here.

This lesson is about the other side: the link someone places without payment and without your asking, because your page was useful to them.

Why most link building happens on your own site

A link is a citation. Someone gets stuck mid-article needing to support a claim, hand over a number, or send the reader somewhere that finishes the job, and instead of writing all that again they link. Which means the question "how do I get links" is really: what is on my site that would save another writer work?

Four shapes of that thing actually work in a small market. A number you measured yourself that exists nowhere else. A tool that takes one specific job to completion. A reference that has not been written in your language, so people are forced to cite an English source. And real news from inside your own industry, the kind only someone doing the work would know.

The cheapest of the four, by a distance, is the measured number, because almost nobody measures anything and publishes it. On our own Linux hosting page we print the search volumes we measured ourselves, with the method next to them and with the limitation stated: that provider has no country filter for Persian, so the figure is worldwide. A page shaped like that is still citable years later.

And here is the boundary of the advice: in some fields nobody links to anyone. If your competitors link out to nothing and there is no trade press in your niche, a linkable asset may earn zero links in a year. That page still pays for itself, but for a different reason: it answers a real search intent. The link is a side effect.

How do you tell whether a page is worth a link from?

Three things make a link worth having, and all three are visible from outside: the host page itself ranks and gets real traffic for something, its outbound links are few, and its subject has some relation to yours. The first and third are quick to check. The second is where everyone goes wrong.

"How many outbound links does this page have" looks like a simple number, but it can be counted three ways and each gives a different answer. On one of our own posts, counted on 5 September 2026: the whole page carries 193 links, inside the article tag the number falls to 18, and if you count only the links inside body paragraphs and drop the internal ones, 3 outbound links remain, two of them sources and one sitting in the author box.

So the same page either "has 193 links" or "has 3 links", depending on how you count. A seller who tells you "our page only has a few outbound links" has almost never said which method they used. Count it yourself:

curl -s -A "Mozilla/5.0" "https://example.com/page/" > p.html
grep -oE 'href="https?://[^"]+"' p.html | wc -l

The first number is always large because it counts the menu, the footer and the mobile navigation too. The number that matters is the links inside the body paragraphs. If those two figures are ten times apart on one page, that page is a directory more than it is an article.

One thing cannot be seen from outside and is better left unclaimed: exactly how much weight Google gives that link. The three criteria above are working experience, not a published Google formula. We do not promise a ranking in exchange for a link count either.

What does digital PR look like without a budget?

Digital PR is a big name for a simple job: telling someone who writes about your subject that you have something useful to them. What separates it from the link-begging emails everyone receives comes down to one thing. Your message has to carry a fact, not a request.

The message that gets answered usually looks like this: it points at one specific sentence in their own article, says you have a number or a reference that completes that sentence, gives the link, and stops. No compliments about their site, no "we have written a very comprehensive article", no offer to exchange. If you cannot find that sentence, this person is not your audience. Removing people from the list matters more than writing a prettier message.

A few real destinations tend to get missed: the trade press of your own industry, the resource pages of universities and professional associations, the blogs of companies that complement you rather than compete with you, and writers who are right now citing an English source because no equivalent existed in your language. That last group is your best chance.

And the success rate? We are not quoting a number, because we have not measured one. What we can say is that most messages go unanswered, and that is the normal state of this work rather than a sign the message was bad.

Where to get listed, and which directories to skip

A citation is any place outside your site where your business name, address and phone number appear, with or without a link. For a local business this is worth doing, but not in the shape sold as "listed in 200 directories". Remember that Google puts "low-quality directory links" in the same link spam list.

The test is simple: does anyone other than an SEO ever open that site? A chamber of commerce listing, the specialist reference for your industry, a municipal site, the site of a trade fair you actually had a stand at, all of those mean something. A site that is nothing but a list of sites, with no users, means nothing, and being listed on it adds nothing to you.

One point matters more in a small market than in English: writing the name and address identically everywhere. "Alpha Company", "Alpha" and "Alpha Co." are three different things, and so is a phone number written once with a leading zero and once with a country code. Making those consistent is a job you can do yourself in one afternoon, and the local SEO lesson shows where it actually fits.

How to mark your own outbound links

Half of link work is the outbound side, and it is usually forgotten. Google defines three values for the rel attribute and writes that links carrying them will generally not be followed:

This matters because "unnatural links from your site" is one of the manual actions Google reports in Search Console, sitting right beside "unnatural links to your site". Selling a credit-passing link on your own site has consequences for you, not only for the buyer.

Our position here is blunt: if you sell ad space on your site, mark it at the moment you sell it, and be honest in the pricing too. What you are selling is being seen.