Using the age of a site as a ranking factor

[aristotle]

---

System: The following messages were cut out of thread Google Updates and Who they Target [webmasterworld.com]

---

There have been many updates over the years, both general updates and targeted updates aimed at specific types of sites. Most likely every new update knocks some sites into permanent oblivion.

This is why I mentioned old sites earlier in this thread. These old sites, say at least ten years old, have been through numerous updates of all types and survived them. New sites haven't been so thoroughly tested.

I already pointed out the staying power of old sites and the long-term commitment of their owners.

So I would like to ask if there been any patents or research papers on using the age of a site as a ranking factor?

[edited by: goodroi at 9:43 pm (utc) on Oct 24, 2017]

[martinibuster]

I would like to ask if there been any patents or research papers on using the age of a site as a ranking factor?

I haven't seen anything like that in a long long time.

[goodroi]

Google has a patent that deals with historical data aka old sites. Here is a good refresher on that patent
[seobythesea.com...]

[aristotle]

Thanks goodroi -- That's an interesting article and I agree with most of it. Although it seemed to go back and forth between "old site" and "old content". Of course there are many old sites, such as old news sites, that have been around for years yet are updated with fresh content everyday. So for just one example, should google's algorithm TEND to give more trust and authority to old news sites over new news sites?

[keyplyr]

We used to talk about the age of the domain all the time. Never really knew if it was a verified ranking factor, but people used to think so. My personal site is 20 years old on its domain and 3 years earlier on a different domain. Still doing extremely well.

[Shaddows]

We used to talk about the age of the domain all the time.

I'm pretty sure age, per se, was an old-paradigm factor, linked to Trust and Authority.

Our relatively old site does fine, but does not seem to have an advantage or disadvantage compared to relative newcomers, once you control for correlated factors like link acquisition (you get more links over time).

That said, I tend to note that older competitors have more stable rankings than newbies, even if that stable rank is lower than the flavour-of-the-month newbie.

[EditorialGuy]

Remember the "correlation isn't causation" argument.

Even if you could prove that older sites tended to do better in the rankings than newer sites do, it wouldn't necessarily mean that age was a ranking factor. It could have to do with inbound links, trust scores, and any number of other things that benefit from years in the marketplace..

[Shaddows]

Remember the "correlation isn't causation" argument.

Even if you could prove that older sites tended to do better in the rankings than newer sites do, it wouldn't necessarily mean that age was a ranking factor. It could have to do with inbound links, trust scores, and any number of other things that benefit from years in the marketplace.

That is exactly what I was saying. We have an old site, it does well. But once you take into account all the stuff that correlates with age, it leaves age itself as a non-factor.

Incidentally, you could make the opposite case to aristotle; that age should be a mildly negative ranking factor. That age gives such an advantage in terms of link acquisition, name recognition and other naturally occurring factors, a failure to acquire these normal factors should be viewed as a signal of distrust. I am not making such a case, just suggesting that age itself is not a clear-cut candidate for being a factor.

[martinibuster]

That article, published in 2011, is a good case in point confirming my comment that patents and research about "age" as a spam catching factor are very old. The patent it discusses was filed way back in 2003. That might as well be the ice age considering all the changes in search technologies that have happened since then.

Interestingly the article states that SEOs came up with the theory that length of domain registration was a spam factor and that Matt Cutts "...went on to rebuff that assertion more than once since the patent was published..."

If you're one of the people who still consider age of a site as a ranking factor, consider this:
Age does not factor into sites caught by Panda and Penguin. They all get swept up, the old and the new.

To use age as a ranking or spam factor is redundant. The definition of redundant is that something is no longer needed.

Age as a ranking factor is one of those easy to understand concepts that "kind of" make sense. But like all things that have historically made sense (like the earth is flat or that Google prefers brands), the sensibility of a matter should not be the deciding factor.

Look for facts that lend credibility and make a theory plausible. Scientific research and patents are the facts that make a hypothesis plausible. Age of a site as a ranking factor is a great example of a commonly accepted factor that (afaik) has no contemporary application in modern information retrieval.

[aristotle]

Of course the algorithm works with many factors such as backlinks, and it's impossible to study one factor in isolation. So let me try to explain what I have in mind with a simple example:

Suppose you've been doing business with someone for many years and found them to always be reliable and trustworthy. Then one day you get a call from someone you've never heard of who is offering the same service. For all you know, they could be a con artist or fly-by-night operator. Maybe you agree to a small trasaction just to test them, but you probably wouldn't give them a big order or job.

Similarly, google has been sending traffic to old sites for many years, in most cases adding up to millions of visitors. Meanwhile these sites have survived doesns of algorithmic updates of all types. They've essentially proven that they're not spammy "churn and burn" types fo sites, whereas a new site still hasn't proven this.

All of my sites except one were launched at least ten years ago. And all of them gained traffic rather slowly at first. But the gains were steady and gradually added up to good traffic. Two of them were hit by the original Penguin update in April 2012, but both later recovered.

So my observation is that it takes time to build trust in the eyes of google's algorithm.

[goodroi]

We should be careful and not make bad assumptions. For example it is not wise to assume that because a patent is old that it is no longer relevant or utilized. For example Google's Pagerank patent goes back to 2001. Google is most definitely still incorporating the PageRank concept into their rankings even if it is not the all powerful ranking influence like it was in the past.

I would also suggest people think of "age" via many different measurements. The age of a domain registration is probably not a very helpful indicator when separating good sites from off topic/spammy sites. However the age of a backlink could definitely be useful to identify quality editorial backlinks. I have seen and researched Google decreasing the value of backlinks when they are modified. That behavior of valuing links differently based on their age would definitely be helpful in Google's fight against paid link spam. Google is on record as not liking paid links and one tell tale signal of paid link is the constant turnover of anchor text & urls.

From my own experience and research I have come across different ways that Google rewards & penalizes based on age. It is not always a direct measurement but sometimes indirect. Just because you haven't been able to notice something, does not mean Google ignores something as a ranking factor. The SEO world is a big place and there are many different situations. Be careful not to blindly follow false assumptions and do your best to test different ideas.

[martinibuster]

We should be careful and not make bad assumptions. For example it is not wise to assume that because a patent is old that it is no longer relevant or utilized.

The point of my comment on site age as a ranking factor is that the research is not current or ongoing. The research was a thing of the past.

The research related to domain age pre-dates improvements such as

• natural language processing
  
• artificial intelligence
  
• machine learning

The fact that it's not mentioned in current research may indicate that it's no longer a viable factor, that it has been superseded. If I am asked for my opinion on domain age as a ranking factor, I would say that site age is very likely not being used.

Here's an example of an algorithm that could be superseded by newer algorithms. Bill Slawski wrote about a link ranking algorithm [seobythesea.com] published by Microsoft that used the domain age of inlinks as a way to give more or less ranking power to the site they were linking to. Thus, newer domains had less ability to confer link popularity to sites that they linked to.

In newer algorithms, there are examples of seed sets based on niches. By default these authoritative sites are going to be older because it takes time to build up the kind of inbound links within their niches in order to achieve a high enough rank to become the seeds. I discuss these kinds of Penguin-like algos here. [searchenginejournal.com] So in those kinds of algorithms, domain age becomes moot. To someone looking in from the outside it may look like older domains have an advantage but what's really going on is something else. To use domain age in this kind of algorithm is redundant.

Thus my opinion that, based on the lack of current research, and the likelihood that domain age as a ranking factor is probably made redundant by the latest algorithms, it's likely that it's not being used.

---

=================== Official Word from Google ===================
I just did a search for any official word from Google and it appears that John Mueller confirmed this year that domain age is not a ranking factor. Thus, my reasoning is sound and confirmed by an official statement from Google.

Google: Domain Age Doesn't Matter For Search Rankings [seroundtable.com]

Case closed?

;)

[aristotle]

The fact is that virtually nobody outside of google knows for sure if age is a ranking factor or not.

When I originally brought this up in another thread, I said that in my opinion age "should" be a ranking factor. That is still my opinion. In this thread, I'm not saying that it actually is or isn't, only that I believe that there's a good possibility that it is.

I prefer to go by what I actually see in the results, rather than by patents, research papers, or google pronouncements. But as I said, I don't think any of us know the answer for sure.

[goodroi]

@aristotle You are correct that none of us can know 100% what is or is not in Google's secret formula. It is very difficult if not impossible to precisely isolate one of the many different ranking signals in that secret formula so we each need to decide for ourselves what we think is in the secret formula.

I would suggest webmasters be careful with their nomenclature. For example "site age" is not the same as "domain age". Domain age usually refers to when a domain was registered. Site age is a more general term that can be defined in different ways (eg when domain was registered, or when site was added to a server, or when content was published, or when the public was allowed access to the site, etc). It is probably a wise idea if the terms "site age" & "domain age" are used separately or it can result in confusion and mistakes.

As I stated in my earlier post, it would be wise for webmasters to look at "site age" with a wider view than just "domain age". The age of your content, backlinks and other factors can have positive or negatives impacts on your performance. For example, I bet most webmasters would agree that freshly updated content tends to have an easier time in Google.

I would beware any SEO that implies age has zero impact on site performance. I agree that domain registration age is not important but to think Google never cares about the age of any site element is a risky opinion IMHO. I would advocate for a more nuanced perspective on site age.

[buckworks]

My domain's age is what it is and it's not something I could change. So wondering whether domain age might be a ranking factor is a moot point, of no practical significance.

It's more productive to think about things I -can- do something about.

[martinibuster]

For example "site age" is not the same as "domain age". Domain age usually refers to when a domain was registered.

Domain age is often used interchangeably with site age and in practice can mean the same thing. The phrase domain registration date is unambiguous.

As for Aristotle doubting something Mueller would affirm or deny, I can understand it if one expresses doubt as to anything John Mueller might say.

I tend to believe Mueller understands what he's talking about, but I don't wish to infer that he is infallible, hence my question mark when I stated, "Cased closed?" The answer may be yes or no.

Let's Muddy Up the Topic!
As for the age of a document, that is, a web page, that can actually play a role in ranking when it comes to a query that is topical and when freshness is important. So it's rather the reverse, a new web page can outrank an old web page, in certain situations.

Back to the Original Post
Sites that weather through update after update I think are like a well manufactured product. Perhaps they satisfy people and people ask for it by name when making a query (sending a signal that this site satisfies users for that query). Perhaps it's longevity results in many citations across a wide variety of web pages from a great many unique domains. Something like that may confer authority.

Joost de Valk, the founder of Yoast told me that he believes it's best to have a product that actually works in order to begin the journey toward ranking number one, that all the other SEO stuff nowadays is almost not necessary. And I can see his point and how a site with solid content, solid user experience, solid service or whatever can acquire all the various signals that tells Google that people like that site and whatever it is they offer.

In that case, is age the ranking factor Google is using or is time simply what allowed the site to telegraph it's appeal and build all those signals?

[Shepherd]

Quoting another member: "google does not want to make sites popular, they want to rank popular sites".

A surefire recipe for becoming popular is to be around for a while providing quality service/products. Is "age" a ranking factor... I would say no. One could have put up a website 20 years ago, and kept is live all that time for less than $1000, that would make it aged but most likely not popular.

That said, in the interest of Muddying up the water: QDF, why did they have to create an "add on" to the algo to make sure "new" content outranked "old" content when query deserves freshness?

[martinibuster]

why did they have to create an "add on" to the algo to make sure "new" content outranked "old" content when query deserves freshness?

I believe it's because the core ranking algo is where you'll find the 200+ ranking factors. But sites ranked in QDF don't necessarily fit into the 200+ ranking factors paradigm.

[Shaddows]

QDF - I've always conceptualised that as overcoming link-weight bias, which would otherwise favour age.

I actually think that QDF is also an old-paradigm factor, superseded by RankBrain (though used to seed that ML process in the first place).

QDF was pre-Caffeine, and therefore could not possibly be part of the modern paradigm.

Here's what I believe to be the original source:

Freshness, which describes how many recently created or changed pages are included in a search result, is at the center of a constant debate in search: Is it better to provide new information or to display pages that have stood the test of time and are more likely to be of higher quality? Until now, Google has preferred pages old enough to attract others to link to them...

He then unveiled his team’s solution: a mathematical model that tries to determine when users want new information and when they don’t. (And yes, like all Google initiatives, it had a name: QDF, for “query deserves freshness.”)
New York Times (June 2007) [nytimes.com]

Reading that article in full, I realise what a sea-change there has been with algo transparency. Different world.

[seoskunk]

google does not want to make sites popular, they want to rank popular sites

I believe that was Rae Hoffman aka SugarRae that said that

Domain Age Doesn't Matter For Search Rankings

Google are very good at answering the exact question. However they don't answer the theme of the question. Yes domain age plays no part in ranking, but links from aged pages does! Its as simple as that folks, now I am going to crawl back under my rock for a while....