A colleague asked me about legitimate link building. He wanted results oriented strategies, something that can scale but wasn’t blackhat or unrealistic “build content and they will come” advice. After thinking about his question that afternoon I realized that maybe links matter less (or maybe differently) for SEO than they have in the past.
Googler Confirms Links Don’t Always Help
In the recent Reddit AMA, Google’s John Mueller stated flat out that links don’t always help and that it’s not the most efficient way to promote a site.
“SEOs often focus on links too much, it doesn’t always helps their sites in the long run or is probably not the most efficient thing to do to promote a website.”
Links Don’t Count the Same as they Used to
These are my observations about the state of links today, with links and citations to various research and patents. It feels like there are at least three ways links affect rankings.
1. Validating Links
Some links are good for getting your site crawled and pass a little bit of PageRank. Obtain enough of these, and like four quarters, you’ll have a dollar. In other words, these links tend to validate that your site is not one of the bad actors on the web.
These are what I call Links for Inclusion and Validating Links. I call them links for inclusion because in my opinion, based on my Penguin algorithm research, Google is creating at least two link graphs. One that excludes pages that are spammy and one that includes non-spammy pages.
This creates a situation where you must have links from the right sites in order to be considered for ranking. This is something called the Reduced Link Graph. It’s called the reduced link graph because it’s not a complete map of every page on the Internet. The map is excluding sites that are bad actors. The map of the Internet is reduced to sites that matter.
There’s a Google patent from 2006 that mentions a reduced link graph as part of a link analysis algorithm.
Here are the most interesting bits from that patent application:
“In a variation on this embodiment, the links associated with the computed shortest distances constitute a reduced link-graph.”
“A Reduced Link-Graph
Note that the links participating in the k shortest paths from the seeds to the pages constitute a sub-graph that includes all the links that are “flow” ranked from the seeds.”
Catch that bit about “flow ranked?” Those are links from high quality sites, that are linking to topically relevant pages. That’s what I was referring to when I said there are links that help for ranking and inclusion and links that don’t, putting you outside of the reduced link graph.
2. Links that Help a Site Rank
If we accept the premise of a reduced link graph, where the web is mapped according to links between relevant pages (or relevant sections of pages), then it becomes clear that some links are going to help a site rank while other links will help it get crawled (validating link) and be included for consideration for ranking. And also, that some links just don’t help.
There is a theory called Site Authority that holds that some sites have a quality called Authority and that obtaining links from those sites will help a site rank. But, Google is on record denying that a domain authority metric exists.
Let’s dismiss the idea of an authority metric and proceed as if Googler’s have been telling the truth that domain authority is not a real metric. The takeaway here is that the idea that news sites and sites with a lot of links don’t really help ranking. And there is proof that this is true.
Forbes, Huffington Post and many other sites slapped no-follows on their outbound links in the summer of 2017 and the SEO community didn’t feel an impact on ranking. Nothing changed because the links weren’t influencing the rankings.
So that’s the essence of point 2. Some links help ranking, some links do nothing, and some links as in point 1, will include your pages in the reduced link graph and put them into contention for ranking.
3. SERPs are Not a List Ordered by Links
The third way links affect rankings is in how they do not affect rankings.
Google is really a popularity engine. The pages that satisfy the most users are what Google will rank at the top. That’s called matching web pages to user intent.
User intent means what the person typing the search query is asking when they type that query and what they expect to see.
What’s important to understand here is that Google is not ranking sites (and personalizing) for the individual user, but for the most users. So when a user types Jaguar, if most users expect web pages about Jaguar the automobile then that’s what Google will rank first.
And it could be that there are more links to sites about Jaguar the animal or more links from news sites to Jaguar the college football team. The importance of this insight is that user intent will beat the link signal.
Links no longer determine who rules the top of the SERPs.
Sometimes you can build all the links possible, outstrip everyone else, and still not rank. The reason you won’t rank is because ranking is no longer an arms race of who has the most or best links.
Link Building May Need to be Reconsidered
So maybe the link building process has to be thought of in a different light. Consider this timeline of link building:
- Ranking used to be about obtaining the most links.
- Ranking became about obtaining the most links with the most keyword relevant anchor text.
- Ranking became about obtaining the most links from the most relevant pages that did not have unnatural outbound linking patterns.
- Ranking became about obtaining the most links from the most relevant web pages and anchor text.
- Anchor text mattered less.
- Content on the page that is linking and content on the page being linked to matters more.
Link Building in 2018?
My colleague complained that link building information found online was unreliable.
“Everything I find online is either 2006 blackhat stuff or unhelpful “build good content and they will come” stuff.”
History of How/Why Links Got Whittled Away
The old link building strategies were built on tactics that initially were acceptable to Google and helped sites rank. The problem was when SEOs scaled them.
For example, submitting to a couple directories was fine. Then SEOs started building directories and pumping up the PageRank score (visible on Google’s toolbar) by buying links.
Another example is trading links. That was the original way of gaining visibility online (before links were a commodity). Then SEOs began scaling it up, sending massive amounts of emails for trading links and Google put an end to that by mapping out what’s natural and what’s not then penalizing unnatural link patterns.
The same thing happened to guest posting. SEOs started scaling it up and ruined that as well. Guest posting used to be a legitimate way to build relationships with news organizations and blogs. Less so today.
Do Links Matter Less?
I think that link building today is past it’s moment of dominance. Links are still important. But they are less important than they have been in the past.
So, rather than spin your wheels worrying about anchor text as if it’s 2006, maybe it’s more productive to focus on where your users are and acquiring links or unlinked citations from those pages.
Any web page that is likely to buy your product or need your service is a good page for your site to be recommended from.
I believe that some people will cite competitors who have reciprocal links, paid links, and so on as proof that Google is ranking those sites because of those tactics.
But what may be overlooked is that those links might not be what’s powering the rankings and that something else is causing those sites to rank, despite those spammy looking links.
Remember, Google is discounting links, not penalizing sites because of those links. That way, the bad links don’t help but whatever good there is about the site will help the page rank.
That’s just one of the reasons why I’m of the opinion that links are mattering in a different way. Links still matter, but not in the same way as they have in the past.