The order that pages appear in the results of a search at a search engine may be influenced by the number of pages that link to that page, and by rankings of the pages that link to that page.
When a site is linked to by a popular and trusted domain, that link might provide more value (and a higher ranking) than a link from a site that is less popular and trusted.
Ages of Linking Domains
A new patent application from Microsoft adds another twist, by also ranking domains based upon domains ages of the domains which link to those domains.
Why?
The cost of purchasing a domain has decreased significantly in recent years, and some domain registrars have offered free domain registrations for up to thirty to sixty-day trial periods.
A spammer might take advantage of an offer like that to build something known as a link farm, which is a spam technique in which spammers “purchase or otherwise obtain a large number of sites and interlink the sites together to increase the sites’ rankings by artificially increasing the number of contributing domains for some or all of the sites.”
The Microsoft patent application is:
Ranking Domains Using Domain Maturity
Invented by Janine Crumb, Krishna C Gade, Rangan Majumder, Vishnu Challam
Assigned to Microsoft
US Patent Application 20080086467
Published April 10, 2008
Filed October 10, 2006
Abstract
Ranking domains for search engines is provided herein. To rank a domain, contributing domains associated with the domain are identified. Additionally, the maturity of each of the contributing domains is determined.
A rank for the domain is then determined based at least in part on the maturity of each of the contributing domains. The domain rankings may then be used to order results for search queries.
This patent application assumes that newer domains have a “higher likelihood of being spam and/or being a part of a web farm that attempts to artificially inflate domain rankings for domains in the web farm.”
By looking at domain ages of sites that link to those newer domains when determining a rank for a domain, domains which have links from older domains “may be ranked higher than spam domains and/or less relevant domains.”
Maturity and Immaturity of Contributing Domains
A search engine may access domain information by communicating with the web servers that those are hosted upon, to access and/or update domain information, such as domain registration date, domain expiration date, domain swapping date(s), and a set of linked domains.
The maturity of a contributing domain may be based upon when that domain was registered or was first discovered by a search engine (if the domain information doesn’t provide a registration date).
Maturity may mean labeling a domain as mature or immature. For example, contributing domains registered more than a year ago could be considered mature domains.
Ranking based upon the age of the contributing domain could involve looking at:
1) Mature Domains only — A domain’s rank might be calculated based in part on only mature contributing domains that are associated with the domain.
2) Mature and Immature Domains — rankings might be influenced by both mature and immature domains, but the value of the rank for the immature domains might be based upon the ranks of the mature domains linking to those immature domains.
While some new domains can be spam, not all are. New domains that are popular, provide value, and gain links from older domains could be allowed to pass along the rankings from the mature domains associated with those new domains.
3) Instead of distinguishing between domains linking to a domain as either a mature or immature, the domain ages of contributing (linking) domain might be used to provide a percentage of ranking to a domain:
For example, in an embodiment, domains that have been registered for more than ten years may contribute 100% of their accumulated ranks to a target domain’s rank;
domains that have been registered from six to ten years may contribute 75% of their accumulated ranks to a target domain’s rank;
domains that have been registered from three to six years may contribute 50% of their accumulated ranks to a target domain’s rank;
domains that have been registered for one to three years may contribute 25% of their accumulated ranks to a target domain’s rank; and
domains that have been registered for less than one year may only contribute 10% of their accumulated ranks.
Resetting Maturity for Expired or Swapped Domains
The maturity of a domain might be reset if the domain expires or if the domain is swapped.
It’s possible for spammers to buy a block of domains that have expired as well as new domains to form a Web Farm. By a search engine resetting the maturity of a domain, spammers don’t benefit from the purchase or swapping of an older domain.
Conclusion
The effect of a process like this based on domain ages might make it look like new domains are being penalized by search engines because they are new (what someone might perhaps call something like a “sandbox” effect).
If a process like this were in place, it might cause new domains that aren’t linked to by older domains to not rank highly, at least until they get some links from older domains.
Do you believe that domain ages in links to a domain should matter in how pages from a site rank?
Hi Bill,
if you wouldn’t have said “Microsoft” then I would have said that this is old news. Google seems to be using the domain age for their ranking for a while already, which also seems to cause the notorious “sandbox effect” that you mentioned.
Microsoft is catching up, that is what it tells me and it raised the question, if Google filed a similar patent for the same thing in the past or not.
If they didn’t and use the methods outlined in this MS patent in their algorithm, wouldn’t they have to pay royalties to Microsoft, because their infringe on their patent?
That’s why do I not like software patents or patents that are similar to software patents in nature, basically logical stuff, where you automatically have to end up at one point in time, if you are trying to get a virtually identical result as everybody else in an environment that is also the same for everybody.
The person who figured it out first registers the find as a patent and everybody else is screwed afterwards, especially the folks who ended up figuring the same thing out (as they logically must have anyway), use it and then get sued, because they did not check the patent DB with millions of crap patents for ones that describe what they just figured out by themselves.
It definitely affects things – but that’s a very arbitrary and rigid scale, and open to the exploits that you’ve mentioned. I wonder if anyone’s done work into an algorithmic measure of domain age?
Also, different ccTLDs have different minimum purchase lengths, and different prices – should those be taken into account when weighing a domain’s influence? Is it even fair to assume that sites hosted on $0.49 .info domains are lower quality?
“Why?
The cost of purchasing a domain has decreased significantly in recent years,”
Because it’s business and they want to earn more… Not because of spammers…
I do absolutely agree to the opening statement that the domain age has a significant impact on the relevance of a website and therefore it affects the organic listing.
The question must be if there is a factor which checks the domain age and the date in time the content was updated. If an old domain is not updated in a certain period the domain will be downgraded and newer domains with up to date content have a chance to climb up the organic index. Otherwise the entire index would be crowded with old domains with relevant but outdated content 🙂
Cheers Daniel
I am aware that Google uses Domain age, interesting that Microsoft filed a similar patent. Thanks for the heads up!
Lets see how we could “break” this part of the algorithm. First of all, one should register a domain name for 10 years, that way you can credit for being registered for a longer period. One builds up content on that domain. After a significant number of articles one starts “aged links campaign” by posting comments on old well established blogs and website in the industry related to his own.
In a short period one will get enough Trust from search engines and his website will start to rank better and better.
What will change? Almost nothing accept older domain with will have even more influence than they do now.
But maybe I’m wrong 🙂
Sasha T.
Bill,
Carsten made several points I was thinking of, about Google’s infamous sandbox, and why now for MS.
Now as I re-read this, do I understand this is mainly or entirely for age of domains with links TO your site, not for the age of your site, really? And if that’s the case, isn’t this going to be kind of far down on the list of things that affect rank, anyway, more like a tweak than a major change?
Regards,
Kelly
Very interesting subject…
We have a site that we are optimising, its 6 year old domain that’s never been optimized and its a hard slog. Yet the top 4 domains for the keywords that we are targeting are 6 months > 18 months old, but all have huge amounts of backlinks, the top position have 48,000 . So, does the age really matter ?
Scott
Hi Vygantas,
The statement in the patent application about the lower cost of domain names points out the idea that it is cheaper for people to buy domains these days to use to create link farms.
While it is good for registrars of domain names that there’s such a demand, it also makes it cheaper for people who want to spam search engines by buying lots of domain names, and pointing links to specific sites from the many domain names they are purchasing.
This patent is a reaction to that kind of spamming, because it means that just buying a new domain, and pointing links from it to another domain will have less impact in the search rankings of phrases found on the pages of newer domains.
Hi Carsten,
It does seem like Google may just be using something like this, and may have been for a few years. I can’t say that I’ve seen anything officially or unofficially from Google that they might consider in rankings for a page, a look at the ages of sites (or pages) pointing to that page. The Google patent on the use of historical data does say something about looking at the age of links, rather than the domains those links come from. It also mentions looking at the length of the registration of a domain.
It’s hard to say whether or not there would be a problem if Google is looking at the age of domains doing linking, or if Google and Microsoft would engage in a patent infringement suit. I do agree that can be a problem with a number of search related patent filings.
Hi SEO Ranter,
I’d really like seeing the kind of study that you suggest, on domain ages.
The patent application describes three different methods that they could use to have the age of linking domains influence rankings. Like most patent applications, there’s the possibility that they could use other methods too, that are reasonably related. You don’t need to describe the process that you are patenting in excrutiating detail to have it granted – just an example of how it could work. It’s possible that something less rigid could be used.
Interesting point about the prices of different ccTLDs. The purchase cost of a domain was brought up to describe why there was a need to add something like this process – because it’s less expensive to buy domains, it’s cheaper for people to purchase them so that they can link to a domain that they want to boost in rankings. That seems to be more of a justification for using domain age than a part of the process. But it’s something to think about. 🙂
Hi Scott,
Under the patent application (we don’t know for certain that it is being used by any of the search engines), it’s not the age of the domain that is important.
At the SES New York 2005, in a “meet the crawlers” question and answer session , one of the search engineers was asked a question about the Google “sandbox.” The sandbox was a place where many SEOs and other people speculated that new websites disappeared into for a certain amount of time just because they were new.
The engineer pointed out a Tsunami Blog (this one, I believe: http://tsunamihelp.blogspot.com/ ) that was extremely informative and helpful after an Earthquake in the Indian Ocean caused an incredible amount of harm. It began to rank for Tsunami within a week of its existence, and saw no signs of stopping after a number of months. It also received a good number of links very quickly after starting.
Did it escape a “sandbox” effect because of the number of links it received? Because of the freshness of the information that it contained? Or the topicality of the subject it covered? Or because a bunch of the domains that linked to it had been around for years and years? The maturity of the domains linking to it are as good a reason as any other.
Age may matter – not the age of your site so much as the age of domains linking to yours, or in a variation described in the patent filing, of domains that have older domains pointing to them.
Hi Daniel,
I’m not sure how much we can state with certainty that domain age has an impact upon rankings. But, it helps seeing patent filings like this one, which are directly from the search engines.
It has the possibility of giving us a little more confidence that there can be some value in getting links to our new sites from sites like Yahoo, or DMOZ or other directories and sites that have been around for a while.
Interesting point on comparing the age of a domain with the age of content on those domains. In some instances that might be helpful, especially in areas where the freshness of content can be important. But for other pages that contain information that doesn’t get stale, it might not – for instance, a page that contains some kind of important historical information, from the Declaration of Independence, to a work of fiction or poetry, to the original charter of a corporate endeavor…
Hi Rob,
You’re welcome.
Hi Sasha,
You may just be right. Just to avoid any possible confusion in your hypothetical, it’s not the length of registration into the future that is important, but the actual age of the domain itself, without any expirations or swapping of ownership.
Someone starting a web site should ideally look for links from a variety of sites, including both newer sites and older ones. No telling if Microsoft will actually do this, or if Google or Yahoo are, but it doesn’t hurt to try to get links to your site from a variety of sources, so going after links from places like DMOZ or other long time directories can be a good idea.
But don’t stop from getting links from newer sites, too. At least one variation described of how this would work indicates that newer sites can pass along the “maturity” of other sites that link to them – so a new site that attracts a lot of links from older sites may have more “maturity” than an older site.
Hi Kelly,
Exactly – it involves the age of links pointing to your site, and not the age of your site itself. This process might have a significant impact upon sites that don’t have links to them from older domains, or from domains that aren’t linked to by older domains. We don’t know how large of an impact it might have…
Yeah, I fully agree that the age of the domain impacts the ratings of a site, but that’s only realistic. Are they punishing sites that are newer? Not really, at least that’s not how I look at it. New sites don’t have many viewers, links or much content. These sties need to start growing those, and then they’ll start to get the links. When you do this naturally, then you’ll get those organic rankings, but it’ll take a while. Which, of course, is normal — because this stuff takes time.
Hi Morgan,
I do want to encourage people who have new sites to see if they can get links for older sites, or from newer sites that have a number of links to them from older sites…
A new site can be considered “mature” under one of the variations of this patent application if it has links pointing to it from mature sites. The patent application explains why – my paraphrase:
It can take time for a newer site to start showing up in rankings. If any of the search engines are following a process like the one described in the patent application, it’s in the best interest of site owners to get some links from older domains, like the Yahoo Directory, or DMOZ, or many others.
So the question is.. does buying old deleted / expired domain names keep you out of the sandbox, if there is a sandbox. I just did some research and found some nice old aged domains available. I found one that is 11 years old, listed in dmoz , and has a pr5..
I’m not saying I would buy the domain for the listing in dmoz or the pr, but the age of the domain would be helpful for a new upstart.
Do you think that the older domain names (says 5 years+) already have a head start with google, I’ll tell you why I ask, One of my sites can only be described as cr@p in terms of looks, design and SEO and with very few backlinks but is 8 years old, and none of my competitors can knock me off the no.1 listing for my search terms.
Scott
Scott
Hi Scott,
If the search engines decide that they should reset the maturity level of a domain to zero when that domain expires or is swapped, then buying old deleted or expired domains to use to link to other sites won’t have much impact on the new sites to which they are linked.
Here’s what is said in the (Microsoft) patent application:
That sounds like a mostly reasonable approach, though I’m a little concerned about the “domain swapping” part of the patent application. People buy businesses and their associated web sites all the time, or go through mergers, or change business structures, or the people whose name is on the registration. The sale or purchase of a domain happens for reasons other than an intent to spam.
Hi Bill,
In replying to Sasha you mention that it is the length of past registration(not future) ideally without any transfers or drops. When / if the algorithm gathers data on the age of a domain do you think it accesses that type of information and how important is it? Also if a domain has been registered for a decade but changed hands, never been dropped but never had a site on it is it any better then another site fresh to the index?
Thanks,
Rob
While having domain age in their algorithm is fine are they also checking when that TLD went live? For e.g. oneindia.in competes with many other .com news sites (India). Obviously .com will be older than .in domains as .in became available just few years ago. So it would be good if they can consider when a TLD went alive (for non .com, .net .org domains)
We see many smaller .com sites doing better than our .in domain on Google only because they are a little older than our .in domain (eventhough their content is not that good)
I was pleasantly surprised the thoroughness of this article, and many thanks to the author for shedding light on this subject.
Something like this could really marginalize new sites in a way that is detrimental to the actual end results.
The biggest side-effect this should have is an even greater increase in the prevalence of domain parking, as the practice could expand beyond the ranks of domaining and become standard practice for anyone who develops/published to multiple domains.
Hi Rob,
Some great questions. We aren’t given some specific details on a number of parts of the process involved in this patent application. It’s also important to keep in mind that the processes the patent application describes may or may not actually be implemented by the search engine.
We are told that the determination of the date could be when it was registered, or when the search engine first discovered the domain, but aren’t provided with much in the way of specifics beyond that.
We can really only guess at how a search engine might treat aspects of the methods described in the patent filing. For instance, a domain that was registered ten years ago that never had a site on it might be treated as a new site because it hadn’t been developed in any way.
Hi BG Mahesh,
A very interesting question – I wonder how much consideration the inventors of this patent even gave to different top level domains, if any at all.
I’ve seen many instances where a business will change the domain name of their site because their business name has changed, or because they were trying to brand their business differently. Having to start with a fresh site could cause problems under the process described in this patent filing, especially if the old site had lots of links to it from older domains.
I’ve also seen many sites that changed from using a .com address to a specific country code address because the owners felt that the change would indicate to their audience that the site was aiming its offerings at people who lived in that country.
Keep in mind that the process described in this patent application isn’t looking at the age of a domain itself, but rather the “maturity” of domains linking to it.
And under one possible way that it could (or may have been) implemented, the idea of “maturity” might mean that a new site can be mature if it has links to it from a number of older sites.
Can you really know that for certain? If age is used, it is only one of a good number of signals (hundreds of signals) that a search engine might use to rank results. Many of those signals may not be all that visible or clear, such as the value of links and anchor text pointing to the pages of those sites.
If you believe that the age is the important factor, your solution may be in creating content on your pages that will attract links from older domains.
Hi Peter,
Thank you.
I wondered about that myself, but I don’t quite think that it is true.
The end goal of what a search engine and a search engineer aims at is to provide as many relevant results to searchers queries as they can. That’s true of whether a site is new or old. It would hurt the search engines to ignore new sites that provide exactly the information that a searcher is looking for.
The concern behind this patent filing, is that someone could buy lots of new domains, and interlink them so that one or more of those domains could have lots of links pointed to them and rank highly in search results because of all of the links. This is a practice that some spammers follow, known as a link farm, and is what the search engine wants to avoid.
Even if this patent application didn’t exist, it’s really in the best interest of the owners of new sites to attract or pursue links from sites that are established and have been around for a while. Those could be links from directories like Yahoo or DMOZ, or many other sources.
Don’t be dismayed at the existence of this patent filing, but look at it as an indication that efforts to make your site something that an older site takes notice of may pay off with more attention from a search engine. And as a way of not being lower in search results because link farms are taking up a spot that you could be showing up in on the search results.
Kelly made now me re-think the whole thing. So what we are talking about is only the ability to pass on link juice to another site. This means that a new site is able to receive “love”, but that only its ability to give link power to anybody else.
That would not cause the sandbox effect. It also makes sense to some degree (in theory). If a whole niche is new and no older sites exist, nobody in that niche is in a disadvantage, because nobody can get much authority links to his site, except from outside the niche sites, like newspapers etc, which would make them authoritative among the other sites.
In established niches would be the effect that new sites could not do much and have as everybody else to try to get attention from its peers and inbound links from authority sites. It just cannot return the favor to the same degree, because it is too “young to vote” yet.
Well, the whole thing should already be covered by PageRank, but Microsoft does not use PageRank (doh) in their algorithm. They have probably a much bigger problem with spammers who setup thousands of sites over night (Blogger for example) and point them all to a target site. In a PageRank scenario would it take a lot of sites to make a dent in a competitive niche, but then MS does not use PR. That they seem to have a problem with this approach showed their issues during the Soccer World Cup 2006.
This patent would not 100% solve the problem, but increase the barrier of entry for a spammer who would be forced to have domains sitting dormant for a while to use them in 1-2 or more years for his purposes. Domain age as factor for a sites ability to pass on link juice would be a good deterrent, because the additional cost, logistics and long term planning is probably too much for the bulk of the spammers.
I think it does. How older your page and how older your links is better for rankings.
Hi Bill,
Thanks for getting back to me and validating a few theories of mine. One theory which I believe has some truth to it is that the age has more to do with registration date then date in the index. Most domains used for spam are new ones so a domain that is aged let alone from say 1998 would of been used for spam ages ago if it were it’s intended use. Transferring wouldn’t affect the domain negatively I’d think as it’s a natural thing to occur over a decade. Possible reasons could be transferring to a family member, a business folding and the division of assets etc… Another variable could be whether the contact info has changed since registration or how long since the last change… Not sure if last variables qualify as over analysis but I have a feeling they do.
Just my thoughts on the topic or maybe its my subconscious building an alibi of the perfect scenario for people in the market to buy aged domains and build new sites.
Thanks,
Rob
Hi Carsten
What many people were calling a “sandbox” effect was that new sites were finding it impossible to show up in rankings for the first few months of their existence – except that some sites did. Also, Google’s Matt Cutts claimed that there was no “sandbox” but that there was something that might be considered a sandbox effect that wasn’t a penalty on new sites.
Considering that many new sites often don’t procure links from older domains early on in their existence, a process like the one described here could cause a sandbox effect.
I do like your “too young to vote” statement for new domains. Except that, under one version of this process a new site that has received links from mature sites may have the value of those links count in its own links – making it no longer “too young to vote.”
PageRank probably wouldn’t make a difference – the value of a query independent ranking factor (like pagerank or other static ranking factors) would be less for links from sites that are immature and more for links for sites that are mature (regardless of whether those domains or old or new).
It quite possibly could, though it’s possible that this cost could negatively impact people with new sites who have no intention at all of spamming anyone, which isn’t a good result at all.
Hi Rob,
You’re welcome.
I imagine that it is tempting to consider that circumstance. While you and I both suggest that it’s not unusual for a domain to have its ownership transferred at one point or another, there’s no discussion of those possibilities in the patent application.
Hi Lyod,
This Microsoft patent application doesn’t really address the age of links, though Google’s patent on “information retrieval based on historical data might consider the age of links, and the length of registration of a domain.
Hi Bill, thanks for the article. This tells us Microsoft won’t be as easy to game and spam as it used to. You could rank Sites easily and faster in Microsoft just by having highly optimized pages. As I have heard Google is been doing this for years, the famous “sandbox effect”.
Hi SEO Guide,
This is one effort by Microsoft to make their search results more difficult to game or spam. They’ve come out with some great papers in the past few years that indicate other efforts, and I applaud them for those efforts.
There’s a lot of discussion in the comments above about the “sandbox” effect. We don’t know with any certainty what it is that Google has been doing that causes some new sites to appear as if they’ve been penalized only because they are new, but I do have to say that many of the sites that I’ve seen that people have claimed were the “victims” of the Google sandbox weren’t very search engine friendly, and didn’t have very many links pointing to them.
uhm.. ok for this patent, good idea, good effects.. but I think that the real change for serps is too heavy and sandbox too long in time.
Hi Bill,
I understand the process described by MS but to continue my theory: you could just “give” me your domain without changing whois information 😉 I guess it wont penalize nameserver change…
Oh, I’m not blackhat, I just like to think about search engines.
Sasha T.
Hi Sasha,
It is a possibility that someone could just give someone else a domain without changing whois information, which might keep a search engine from “resetting” the age of a domain under something like this patent filing.
If a number of new links pointing to and from a site start appearing, and the content of pages start changing rapidly, a search engine might take those changes into account, too.
I wrote about a Yahoo patent filing that track those kinds of changes in “Yahoo on Using Exceptional Changes in Snapshots of the Web to Ban, Penalize, or Flag Websites” and Google’s patent on Information retrieval based on historical data also describes following such changes.
So a transfer of a domain could be difficult to spot, but if there is a dramatic repurposing of the content of a site, and the links to and from it, a search engine might notice the changes…
Personally, I can see the purpose of considering the age of the domain for ranking purposes, but I also don’t have to agree with it. There can be 10yr old domains with crappy information and a 1yr old domain with excellent content, and the older domain will rank higher all things being equal? There has to be a better way to gauge the content of a site and the intention of the site owner. If someone thinks of it, let me know so I can apply for the patent… ha!
Nice post and discussion here on domain ages. Feels good to be here.
Hi Addies,
Good to see you here. Welcome.
I don´t know why the age can affect my web site page rank… imagine this… I have lot´s of domains that I registed 5 years ago, but they have nothing… soon or, some had something, but now, I´ll add new content 100% diferent than the other.. I think the link ideia is the best one…
Hi Ruben,
If the domains didn’t have actual sites associated with them, it’s possible that the search engines may not look at the registration date, but rather the date that they first discovered the site, whether through a search crawling program, or a toolbar, or some other way.
If the content of an older site changes completely, that may be a significant signal to search engines that a site has changed in purpose. It’s possible that may also have an impact upon how the search engine might treat links from such a site.
My question is how much it matters that older sites link to -you-?
Hi Andreas,
It appears that the assumption behind this patent filing is that if a newer site is linked to by a number of older sites, that it is less likely to be part of a link farm, and therefore less likely to be spamming a search engine.
I am still waiting for a relatively new website of mine to get a PageRank. I believe Google has something similar to this where they only rank your pages after a few months.
Hi Horsham Web Designer,
Thanks for the question. There are a couple of things worth keeping in mind when starting a new site.
One of them deals with PageRank. If you are trying to keep an eye on the Google Toolbar, to determine if the pages of the site has accumulated PageRank yet, remember that the PageRank that shows up on a toolbar is somewhat different than the PageRank that may play a role in how your pages are ranked at Google.
The PageRank information for a web page in the Google Toolbar is only updated a few times a year, and isn’t a reflection of the actual PageRank of a page, which can be updated much more frequently. So, your new pages may have some PageRank, even though the toolbar indicator doesn’t show that they do.
PageRank for a page is based upon the PageRanks of pages linking to your site, as long as those links are straightforward text links that aren’t using temporary -302 – redirects, or a rel=”nofollow” attribute in the links. As more links like that are pointed to the pages of your site, your pages should start to accumulate PageRank. When you first start out, chances are that you have little or no links pointing to your pages, so you aren’t going to have much in the way of PageRank.
One of the models behind PageRank are citations in papers found in academic journals. When a paper in one of those journals introduces some great ideas, it may be cited by other papers. The more it is cited, and cited by other interesting and important papers, the more important it is. Under PageRank, links are like those citations. The idea is that the more links a page has from other page, the more important it must be. And if a page has links to it from other pages that have been determined to be important, it must be important too.
PageRank isn’t the only ranking factor that Google uses, and it shouldn’t be your only focus. Making sure that you use words on the pages of the site that people searching for it will use to find it is very important as well. Making sure that the site is easy to use can also be very helpful. Those are things to work on in addition to creating content that people will want to link to, and in buildinig links to your sites.
Hi William, firstly thanks for the reply. I don’t actually use the Google toolbar I use a firefox extension but I expect that these show the same value anyway. You mention the PR on the toolbar only updates a few times a year – is this when the ‘Google Dance’ happens?
You’re welcome, Horsham Web Designer,
Some good questions. 🙂
The PageRank values on the FireFox extension do come from the same sources as the PageRank values displayed on the Google Toolbar.
We know that the updates to those values are independent of changes to actual PageRanks for pages.
The “Google Dance” was an actual update to Google’s index, which appeared to happen every four or five weeks, until possibly sometime in 2003 or 2004. A fairly simple but informative description of the Google Dance can be found here:
link no longer available
Google updates its indexes much more frequently now. A couple of nice posts from Google’s Matt Cutts are worth looking at on updates to Google’s index and to the Toolbar PageRank:
Minty Fresh Indexing
More Info on PageRank
Note that the second post tells us that sometimes toolbar PageRank may be updated a little more frequently for some URLs. Regardless of how frequently it is updated, it isn’t the realtime PageRank for a URL.
Without a doubt the age of a page, domain and the age of links to a website are a contributing factor in SERP’s
Hi SEO Norwich,
I’m not completely sure that we can say with certainty how search engines might be using the ages of pages or domains or links, but it does seem reasonable for them to be paying attention to signals like those.
The reason why a domain age affects rankings is because of this presumption that older domains are more valuable compared with a new unproven domain. The logic behind this status quo is the data usage – the site’s history on rendering, over time, useful content for its community.
In addition to domain age there is another trust measurement, namely link age; as with domains, links become more potent as they become older.
For new sites to improve buoyancy in the SERPs and avoid being sandboxed right from their early days, they need to follow a tight content publishing schedule backed by a sustainable link building program.
Great content, as usual on SEObytheSea!
Thanks, Bill!
Cheers,
Cristian
Hi Cristian,
I do think there is an assumption that older domains may be more valuable, but there’s also a very practical aspect to that assumption as well – the search engines actually have some history about a site, about visits to its pages, about the links pointing to it from other places, about how similar it might be to other sites.
We’ve been told by Google that they never set out purposefully to create a sandbox effect, but rather that other things that they have done can keep very new domains from ranking well after possibly initially appearing in search results. Regardless of whether or not they did, creating quality content and attracting and building links on a steady basis can help overcome the impact of such an effect.
Does the search engine look at when the domain first registered or at when its first index in the serps.
Hi Lee,
Good question. I looked through the Microsoft patent, and found the following line:
There are a couple of things to keep in mind, though.
The first is that under this process, a search engine might consider a domain to be “more mature” if a good number of links exist to it from domains that are older.
The second is that a change in the ownership of a domain might set the clock back on how mature the search engine considers the site. So a change in registration, or a complete revamping of the content of a site might cause a search engine to consider a domain to be less mature than it is.
The domain “example.com” was booked just over a month back, and despite a reasonable number of links, including from a PR8 site, has not risen in the search rankings much. All sites in the SERPs that this site targets are 4-8 years old. Age does matter, IMHO.
Subsequent to my post above, as an example, see the ages of all the sites that turn up for “london hotels” on Google (a search that example.com targets). All of them in the top 10 are at least 10 years old. (Install SEOQuake for Firefox to see the ages on the SERP page itself.)
Hi Raghu,
Given the adult nature of the classifieds on the front page of your site, I’ve removed your references and links to it. I think the problem you’ve been experiencing can be discussed without them.
The site has been online for about a month, and you’ve acquired a number of links to it. You’re attempting to rank well for some popular and highly competitive terms, such as “london hotels” against sites that have been working hard for at least 3-4 years, and possibly even much longer, to rank for those terms as well.
I’m not sure that you can just point to the age of your competitors, and the age of your site, and say that is a factor that matters.
Chances are at this point that Google may not have even found all of the links that are pointed to your site yet, and may not have fully indexed the pages of your site. It can take some significant time and effort for a new site on the Web to start being competitive, especially against sites that have been working hard to be competitive themselves for a while.
I have seen a few sites rise very quickly in rankings at the major search engines, but those covered very timely topics, acquired a very large amount of links in a very short time, and provided fairly high quality content.
Hi Bill,
I’m agree. For better ranking, first, CONTENT is the key. and second high quality inbound links from proven websites .edu, .org etc. (with at less PR5).
Hi Juste,
Quality content does help, but I’m not going to ascribe any significant “authority” to .edu or .org domains, or to a page with a baseline amount of PageRank.
I’ve seen significant amounts of link spam on .edu and .org sites that weren’t managed or controlled very well by their owners.
Of course a link from a page that has a high level of PageRank may be helpful, but if Google is following the reasonable surfer approach, it’s possible that you may get more PageRank flowing to your page from a PageRank 4 page than from a PageRank 5 page based upon features associated with the link, the page the link appears upon, and the page the link points to.
Hello Bill,
There has been a lot of talk over this already, but it still makes me wonder what really makes it a factor for rankings? I get the idea that newer domains have more chances of being spam but we can not deny the fact that old domains can’t be turned into spam after some other person buys it.
What I am trying to ask is that, is it really the age of the domain that is a factor or the age of the link profile, the type of links and the age of the links it has? and does an old domain carry more weight simply because it is old or because of its past records (if it has any)?
Hi Akash,
The Microsoft patent from the post looks at the age of the domains that link to a page, and makes an assumption that sites that have been around on the Web longer may be a little more trustworthy than sites that are brand new. While someone might buy one of the older domains, there are signs that a search engine could look at that might indicate that an older site might have changed hands, such as a change to the content of that site, or a change to the kinds of sites that older site is now linking to.
For instance, imagine that someone has been blogging at a site for 5 years, and they tended to link to government sites, wikipedia articles, and other blogs. All of a sudden, the site starts linking to casino sites, weightloss pages, and pharmacy sites. That change is the kind of thing that might trigger a search engine to reevaluate the site, and possibly consider that it might have changed hands, and is no longer quite as trustworthy as it might have been in the past.
The patent notes that in an instance like that, they might “reset” the maturity of a domain.
Bill,
What’s your view on dropped domain names? I registered a domain name recently (to start a lifestyle blog) which was dropped several times. When I did a “site.com” Google Search some p0rn results are showing up. Its captured twice by archive.org but the page cant be viewed. So cant decide the category of that domain. So will it hurts its trust on Google?
– Mahesh