The Importance of Links for SEO

Links pointing to your site are among the strongest signals for Google. If you have lots of trustworthy sites linking to you, then it’s reasonable for Google to trust your site too. That’s Google’s big invention.

In the past, links didn’t really matter for search engine optimization — back then, Altavista was the favorite search engine in Sweden. Altavista more or less only measured how many times a word appeared on the page and ranked it based on that. Google, however, completely changed the playing field when they arrived, and their main weapon against that type of spam was: links.

PageRank changed the map

Google’s PageRank algorithm is a system for measuring links. It measures how many and how important the links pointing to a specific page are (note: page, not site) and evaluates the page based on that. PageRank is a fairly complex algorithm and is constantly being developed, but its function remains the same as always: to make it harder to manipulate results by measuring authority. The big advantage of measuring links rather than just the content on the page is that it’s much harder to manipulate links. This creates inertia in the system — links simply take much more time than just writing “Britney Spears” 200 times on the page, which was quite common in Altavista’s case.

So why does Google still have problems?

Google has problems with its index: there’s a lot of junk, and some of it is link-related. Most of the warnings they send (Google warns in Search Console if they find something they believe breaks their guidelines) are about the type of spam that is done on one’s own site — indexing lots of auto-generated pages (such as search result pages, for example) or working with hidden text (for example hiding an H1 heading from visitors but showing it to Google). The reason for this is probably not that less scrupulous SEOs don’t understand that links are important, but that it takes very few resources to manipulate things on your own site. Take the indexing of search results: it’s even a bit blurry what counts as spam and what is simply good SEO. On this blog, for example, we have category pages — these are really nothing more than a search result, and it’s perfectly reasonable to have categories on a blog. If I were to put all posts in 10,000 different categories, however, so that the majority of pages on the site were just different sorts of the same posts I’ve written, then that would be spam. Somewhere in between, you find the line.

Google works hard to improve its search results, and updates are released continuously. Hundreds of updates are made every year, and today the algorithm is very complex and largely self-learning. That automatically means it’s difficult or impossible to fully grasp, even for those most deeply involved in Google’s machinery. In recent years, Google has realized that they need to work together with the SEO industry to improve things — in contrast to the early years, when they probably felt the opposite.

There is work with links at all conceivable levels, from the worst junk bought on Fiverr to well-deserved links from universities. And it’s precisely how links are measured that has long set Google apart from other players. In recent years, this has started to change: Google measures more and more different factors, and perhaps we’ll see diminishing importance of links in the future. The problem is that it reduces inertia and opens the door to easier manipulation.

Think relevance, all the time

Ten years ago, Google was not nearly as smart a machine as it is today. Back then, it was enough for a really strong site to place a link in the footer and suddenly you’d climb ten positions. That’s not how it works today. Partly because Google’s Penguin filter is now a constant part of the algorithm, and partly because they are much better at recognizing relevance. A relevant link, in a relevant text, to a relevant page is worth so much more than a footer link. It also drastically reduces the risk of filters being triggered.