Lately, we've been seeing a lot of chatter on forums about the age of domains and links. No, we're not talking about the Age of Domains and Links - that time is already upon us (we call it "The Internet")! Instead, note those lowercases - we are talking about the birthdays of domains and links and whether or not age affects a site's rankings or the power of a link. A lot of people think that, like a sagely seaman that acquires wisdom and whiskers as his years trickle away on the sea, domains and links gain value as they age. Sure, age maybe possibly could play a small part, but as Rand will show you in the following video, we're nearly 100% sure it's generally not something you should worry about.

Video Transcription

Hi, everyone. Welcome to a new edition of Whiteboard Friday. Today we are talking about the age of site and the age of links. You might have seen, I've been seeing a lot lately, there's sort of a trend of forum threads popping up, questions popping up, people sort of asking and almost dealing with this idea out in the ecosphere of SEO that an older link, an older domain, and a link that's been around longer all necessarily provide more value. They are better for rankings, are better from a search engine perspective. They get crawled more often, these kinds of things.

I would question this. I would strongly question this. It's not just because Matt Cutts sort of had a video where he talked about this. Someone asked him a question on this topic and he said like, "No, I wouldn't worry about that. It's buried 38 lines down in some patent from 2002, but that doesn't mean it's a ranking factor for us." I've heard him and other Googlers on stage at conferences be even more forceful than he was in that video in saying that, "We can't promise that it's not a factor, but if it is a factor, it's super tiny. It's a really small thing. We're using it primarily in conjunction with other things to try and see what's going on." So I wanted to try and address these critical questions about the age of a site and old links and the age of links.

First off, does the age of your site matter? Now, this is purely from an SEO perspective, in terms of if I have exactly the same effects going on, the same number of links pointing from the same places pointing to the same pages with exactly the same content. But one of those sites was produced in 2008, and one was made in 2000. Is that going to substantively change rankings? The answer is almost certainly not. Almost certainly not.

Now, the thing that you're going to say and that I hear a lot of criticism about when this myth is debunked or when Google says no that's not the case, is people have this idea like, "Well, wait. I look. I go through a set of search results. I see these top ten results. It's '04, '03, 1998. These are old, old links. So, how can you tell me that oldness doesn't matter? That domain age doesn't matter?" The reason is I think what you are observing is correlation not causation. It tends to be the case that people who registered those early domains in a lot of those sectors, they did a lot of things right. They got great domain names. They often got exact match domain names, sometimes short domain names. They've built brands. They've had a longer time to build those brands, which means more people are aware of them, more people know about them, link to them, and reference them. They've been in the media and the press. They get included in directories and lists and linked to by big and important people. The amount of time that a business has been around necessarily impacts all of these reference and citation types of impacts. For that reason, you're going to see that a lot of these earlier sites look like they perform better.

But I would actually say that if you produced a site today and could accomplish all the things that a site that is ten years old has done, you would actually outrank them. The reason is because you've earned those links, that reputation, and that brand faster, and therefore your acceleration rate is much greater.

When people look at something like a Twitter, they say, "Wow. Twitter is this amazing company. In the last four years, they've grown to 120 million users." But they don't compare it against something like eBay which has many hundred millions of more users, but they do it in the context of the time that they've been around and what they've been able to accomplish in that time frame. So I think this is a matter of correlation and not causation. I'd be careful about ascribing pure value or thinking like, "Oh, I'll never be able to outrank those guys because their site was made before mine." Don't worry about that.

The second thing is, what about registration length? Google did issue this patent, the same one that I think Matt Cutts was referencing in his video, where they said, "We might look at how long a domain is registered because it tends to be the case that spammers and domainers and affiliates and people who we think might be manipulating our search results, they often will register the domain for the shortest possible amount of time." There was even, for a while, this idea on domain tasting where you could get a domain for 30 days or something and then give it back. Therefore people who are registering for a year or two years, we might frown upon them compared to the people who registered for five years or ten years. So, there was this rush in the SEO community like, "Oh shoot! Let me go and register my domain for a really long time."

Again, I'm going to call mostly myth on this one. Yes, it could be the case that in conjunction with lots of other signals they go, "Wow, we see this signal that looks like spam and this signal that looks like it's manipulation. It looks like they're getting these shoddy links. Their domain is only registered for a year. They're on dot cc domain or a dot info domain." They can look at metrics like that, and if they pattern match those against spam, maybe it would hurt you, but I really wouldn't worry about this. If registering for five years or ten years is tough on your budget and you want to save that money for something else, go for it. I wouldn't sweat it. If you have the budget though, I probably would register for a longer period of time. Not necessarily for the SEO reasons, but for the reason that you might forget about it next year, and it is really a pain to have to go back and register if you've forgotten or you're on a trip when it expires and you didn't check your e-mail. It's a nightmare. Or they get your e-mail wrong or something. That's problematic. So for that reason, longer domain registration might help.

Third question, big one. Do old links or links from old pages or old sites boost rankings more than new links? There are two ideas inherent in here. One is that here are pages from different time frames. One is from 2000, 2004, 2008, today. One idea is that oh, this link is more powerful because it's on a domain or on a page that's been around for a very long time. The second idea is oh, it is not because it's on a page that has been around a long time, it's because the link itself has been there a long time and that as links age they grow in power. I don't know, like some sort of wizened Jedi over the years. The force grows stronger within him or her.

I wouldn't actually dispute both of those ideas. At least, I would dispute the idea that those are the reasons why they get more powerful. Yes, it is probably the case, again like we talked about, that a lot of the time if a link has been around a long time on a page that's been around a very long time, it's an important page and an important site. A lot of the Web decays. We were looking at decay on the Web in our Linkscape index, and a full 80% of the URLs that we requested and got data for, link data, we found 200 response codes, we found content on those pages, within about a year, between 12 and 18 months, that content's gone. If you think about the fact that there is this huge billions of page index for the World Wide Web but only about 20% year-over-year is really surviving. Yeah, you can kind of understand. Hey, those are the important pages. They keep getting links. They keep getting references. They're clearly still in business. People still think they are important enough to point to. Those might be the reasons that those older links are passing more value. Not just because they're old or because they've been there a long time.

As a corollary to this question, a lot of people worry about and ask about, "Oh, no, what if my link disappears? What if I take down a page and put it back up, or one of my link partners has that page 404 for a few days, or repoints it somewhere and my link is not on the new version and then it reappears. Will I lose that aging process?" Like a fine wine it will suddenly be like somebody popped the cork and put it back in and now it's exposed to the oxygen and won't taste as good? No. No, no, no. I can't tell you how much I would not sweat this. Even if you are sure that I am wrong about this stuff, believe me that this is not going on. It is not the case that if your link was created in 2001 or was on a page from 2001 and then in 2004 for about a year it disappeared, it probably hurt you during that year that it disappeared. When it came back in 2005, for the last six years, it's not like oh, that's so much worse than if it had only been there that one year. You could apply this to days or weeks or months as well. So I would not be stressing about these kinds of things.

What I would be worrying about in terms of links is where does the link come from? Is it a great site? If it is a great site, it is relevant to your audience, it has a lot of traffic, it is sending you good traffic, it has good anchor text, it's pointing to the right pages, and the domain that it comes from is a strong one, the page that it comes from is well linked to, great. Don't stress about putting this age stuff in here. I would bet that if we did a ranking correlation, we would probably see exactly the pattern that a lot of SEOs see, which is old domains appear to rank better than newer ones. But I would disagree that if we were able to put that in the ranking model, which we'll try this year at the ranking factors, put in the ranking model that you would see any measure of causation coming out of that. I think this is something where it is perception only.

All right. With that, I just want to tell you about one more thing. Next Friday is a very exciting day for us. It is Friday January 21st at 10:00 Pacific, that's 1:00 p.m. Eastern time or 6:00 p.m. out in London Greenwich Mean Time, we're going to have our first ever live Whiteboard Friday. That means that I'll be here in front of this and you can watch me live. You can Tweet at me and I'll be answering Tweets directly off our account. That should be tons of fun. I hope you'll join us live. Just go to the blog right around or just before 10:00 a.m. Pacific time. You can join in. We will also be recording it, so you can watch it afterwards if you happen to miss it or you are busy at that time.

Thanks so much. Take care. We'll see you again next week for another edition, a live edition, of Whiteboard Friday.

Video transcription by SpeechPad.com

Mark Your Calendars!

Live Whiteboard Friday next Friday at 10AM PST with live chat!