Reader Q & A: How Link Signals Can Be Misunderstood

Most readers here are familiar with the concept of search engine “signals”, i.e., things the bots look for and count, and algorithm’s use and score, in order produce the search results page.

While all of us agree such signals exist, this doesn’t mean we all agree what those signals are and to what degree any of them matter, can be altered, exploited, or manipulated.

There have been several excellent documents that aggregate opinions on signal factors and values. Here’s one: 30 Link Builders Discuss Backlink Analysis For Campaign Design.

I’ve read that there are hundred of signals, that they change all the time, and that certain signals matter much more than others.  To some extent this is all true, but it leads to great confusion for the general web marketer that doesn’t specialize in link building, and to be honest, it leads to confusion for many of us that do specialize in link building. Get the top link builders in a room together and ask them to discuss anchor text until they all agree, and you might be there a while…

Today in Link Week is our first Reader Q & A.  The below question comes from SEL reader Megan B. who asked:

“Hi there, do you have any articles on whether or not the age of the link matters for SEO (meaning, the longer the link is on the page, the better)?”

Answer

First, this is a fantastic question. Many people assume old links are better and don’t think through multiple scenarios logically. Your question shows you are not willing to just take the standard “old links are better” answer. I’m going to go out on a limb here and invite abuse by actually trying to answer your question. (You can read what follows using the voice of Mr. Miyagi from the Karate Kid, Charlton Heston from the Ten Commandments (aka Linkmoses), or the voice of Daffy Duck. All are applicable.)

Link age alone is meaningless

The rationale behind the question is: do engines like old links better than new links? Or perhaps better said, old links might be perceived as being more trustworthy than new links. On the surface, this seems logical. An older site, especially if it existed before Google did, would be less likely to have links on it that were rankings driven, right? If the links existed before Google existed, such links wouldn’t have been for search rank, because links did not affect rankings yet. Nobody chased links to improve search rank.

Yes, this is completely and unquestionably true. However, not all old links are of high value, because not all old sites are of high value. Many gray-or-darker SEO link builders bought (and still buy) old sites to be in control of the links on those old pages. Age alone does not a quality link make.

There must be additional signals confirming the quality of both the site giving and the site receiving that link.  Think Geocities.com.  One day, thousands of old pages with millions of old links, the next day, gone.  Most were worthless long before they were gone, even though they were old.

That’s just one easy example to illustrate the larger point. Just because a link is old does not mean it has any value whatsoever.  There are plenty of old and abandoned web sites, blogs, and forums out there. The value of such links will depend on who created the content, their original intent, and who linked to that content then and now.

In 1998, I built a site for my wife’s stained glass business. For health reasons, she had to stop making stained glass shortly after the site went live, but the site sat live on the web until just last year, when I finally got around to killing it off. It was a 12 year old site within a vertical that was completely legitimate and it had a vetted links section. It had been crawled, it was old, it was real. It even had a link from Yahoo. You might think it was pure gold from a linking perspective, but is was absolutely worthless because I’d never told anyone it existed, and thus nobody else ever linked to it, other than Yahoo!

Now imagine a Geology professor at MIT creates a web site in 1998 devoted to useful geology links, and for many years maintains it as it becomes a site that attracts links from other geology sites all around the world. Call it an “authority site” if you want to. The professor retires and the site is only updated once a year by grad students. Now that’s a page with some very powerful link signals. My wife’s site and the MIT professor’s site are both 14 years old, both are legitimate, both are about very specific topics, and both link out to other sites.

But they couldn’t be more different in terms of the value of the “old links” within them, because one site became a respected source site within its field, while nobody ever knew the other site existed.

If these two examples represent two extremes, then all web sites exist somewhere between these two extremes. A site can be old and helpful, old and useless, new and helpful, new and useless. You must be able to recognize which is which, and why. This is where long practicing link building strategists can be of help. We have tried many tactics over the years, observed countless algorithm tweaks and updates, and we know it when we see it.

This type of answer tends to frustrate people, because they want a concrete answer on which they can base a strategy. If I answered “it depends”, it would be the correct answer, but not a very helpful answer.

I hope the more detailed above example and the approach I took to explain my beliefs and experience about link age were helpful to you.