Search Engine Watch
SEO News

Go Back   Search Engine Watch Forums > Search Engine Marketing Strategies > Search Engine Optimization > Link Building
FAQ Members List Calendar Forum Search Today's Posts Mark Forums Read

View Poll Results: Which links are more valuable: older, newer or neither depends occasion & location?
Older Links 5 31.25%
Newer Links 1 6.25%
Neither - It depends on the occasion and location. 10 62.50%
Voters: 16. You may not vote on this poll

Reply
 
Thread Tools
Old 03-12-2005   #1
Nacho
 
Nacho's Avatar
 
Join Date: Jun 2004
Location: La Jolla, CA
Posts: 1,382
Nacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to behold
What is the Value in Links - OLD vs. NEW?

Which link do you think is more valuable, older links, newer links or neither because it depends on the location and the occasion?

Here are some examples using searchenginewatch.com to illustrate the differences:

OLDER LINKS
Looking on the left navigation you will find a section called "Other Helpful Forums Related to Search:". Another example we see in the SEW Blog (and most blogs around the web) are the "Other Search Blogs" or the "Other websites we read" links. In most cases, these links have been there for a long time and remain as a vote of confidence. Other examples are links on the articles themselves that point to sources, in which will most likely remain forever.

In a way older links are those that a search engine could identify that the link has remained on a particular page and not moved. More examples of these are deeper pages for: archived articles on news sites, or press releases on websites, products details pages on ecommerce sites, forum threads, etc...

NEWER LINKS
Typical examples of newer links are probably when Danny Sullivan, Chris Sherman or Gary Price write a blog telling the readers about a new technology or perhaps pointing to an article/forum/blog on another website to discuss their point of view on the subject. These blogs when posted make the link visible directly from the SEW Blog homepage. Then these blogs move on from the homepage to being archived and only accesible through the permalink turning now into an "older link".

In a way newer links are those that a search engine could identify that a link is promoted on a page that has continuous fresh content but it will eventually dropped or archived. More examples of these are homepages for: news sites, press releases on websites, ecommerce sites, forums, etc...

Please share what you think should be more valuable and why? You could also dicuss how based on your testing you believe specific search engines view such links and why you believe they do it in that format.

Last edited by Nacho : 03-14-2005 at 07:42 PM. Reason: typo
Nacho is offline   Reply With Quote
Old 03-12-2005   #2
NFFC
"One wants to have, you know, a little class." DianeV
 
Join Date: Jun 2004
Posts: 468
NFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to behold
Oh boy, my favourite subject!


I think this is an excellent topic, it is one that should encourage us all to take a step back from our day to day SEO grunt work and really take a long cold look at things.

To me there are too main schools in SEO, the doers and the dreamers. I'm a dreamer, my SEO involves doing things that *I* think are right and trusting that sooner [I hope] rather than later the search engines view aligns with mine. Doers on the other hand aim at the current bullseye. Not saying that my way is better, just that it works better for me.

>Which link do you think is more valuable, older links, newer links

This is where it becomes interesting to me, the only logical answer is both. This provides a problem for the search engines, many of which have been unable to deal with the issue. I remember a quote that Tedster [of wmw fame] posted regarding Northern Light, their aldo guys were saying in essense that they had a factor that rewarded new stuff and one that rewarded old stuff, even they didn't know how an index would turn out.

I've thought about this a lot, my view is that old domains with fresh content should be "rewarded", this site being a perfect example. Old on its own shouldn't cut it [and I think that Google for one has got this wrong lately] nor for that matter should new.

Holy grail...old established domain with frequent updates in both content and inbounds. New sites should be looking more at where the links are coming from, pagerank is dead but in a way is more alive than ever, all links are not equal.

This is worth a read too, http://jeremy.zawodny.com/blog/archives/000751.html pretty old though
NFFC is offline   Reply With Quote
Old 03-12-2005   #3
randfish
Member
 
Join Date: Sep 2004
Location: Seattle, WA
Posts: 436
randfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to all
There's a good paper on this subject as well from some IBM researchers - Trend Detection through Temporal Link Analysis - http://einat.webir.org/JASIS_2004_Te...s_Analysis.pdf - watch out, it's a PDF.

They note that temporal link data is not being used (this was Jan. 2004) and also say that it can be integrated with static-link measures for greater relevancy. Interestingly, they also suggest measuring traffic levels to websites to further improve the data quality... scary!

Last edited by randfish : 03-13-2005 at 04:02 PM. Reason: year in post was wrong
randfish is offline   Reply With Quote
Old 03-12-2005   #4
Nacho
 
Nacho's Avatar
 
Join Date: Jun 2004
Location: La Jolla, CA
Posts: 1,382
Nacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to behold
Good poins NFFC! As you can see, it's one of my favorite subjects too.

We could of course take this topic to a very technical level and discuss things like the Temporal Link Analysis paper, but I'm focused in finding more logical interpretations which I believe is your approach as well.... right?
Nacho is offline   Reply With Quote
Old 03-12-2005   #5
NFFC
"One wants to have, you know, a little class." DianeV
 
Join Date: Jun 2004
Posts: 468
NFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to beholdNFFC is a splendid one to behold
>more logical interpretations which I believe is your approach as well.... right?

I wouldn't say that, my approach is more an emotional one. I think SEO is an art not a science, the good bits of it anyway. Don't get me wrong a good bit of reading of the old and new research papers is a a positive thing, but ultimatly the www and therefore SEO is about people not formula's.

There are only, in the realm of www IR, so many ways to skin that cat. The old v new conundrum is a great way to focus on that fact imho.
NFFC is offline   Reply With Quote
Old 03-13-2005   #6
carpediem
Member
 
Join Date: Oct 2004
Posts: 5
carpediem is on a distinguished road
>my view is that old domains with fresh content should be "rewarded"

I agree with this general concept, although I really don't think the age of a links have a factor. There is evidence that older sites have an advantage, but this doesn't necessarily relate to older links.

I have always wondered whether the RATE at which links were acquired had an impact. Some say that they can have a negative impact if you acquire them too quickly for a new site. If acquisition rate is actually used, then couldn't the rate of link acquisition also be applied elsewhere. For example, would either of these sites have an advantage:

1) A site that has been around for 6 years and it gets very few new links (but it has a large # of old links from authority sites)

2) A site that has been around for 18 months and is steadily gaining a high # of links from authority sites.

Or what if you have an older site that has a huge amount of inbound links from authority sites, but it is steadily losing links? Could a newer site that is heading in the opposite direction (as far as link acquisition rate) beat it out with far fewer links?
carpediem is offline   Reply With Quote
Old 03-13-2005   #7
Nacho
 
Nacho's Avatar
 
Join Date: Jun 2004
Location: La Jolla, CA
Posts: 1,382
Nacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to behold
Hello Carpediem,

Welcome to SEW Forums! I believe you are looking at only they way links are acquired. However, the intention of this thread is to look at links in the same way a search engine would view them in a logical or "emotional" manner. Think of a crawler coming back to your website over and over again to study your link structure, both internal and your outbound. Forget about your inbound links for a second, so you get an idea of what I mean.

Here is a quick story for illustration purposes:

Say you have an ecommerce website. You place a brand new product that was just launched on your homepage. The crawler comes back, hits your homepage and says "Oh this is a new link. Never seen it before." So, it takes it back for evaluation. It comes back to your website and now it finds the same link in the homepage and within a navigational category. This time the crawler says, "Oh, here is that link again. Seen it before on the homepage." Then you think that by now your new product's page has been indexed so there is no reason why to keep it on the homepage forever because you've got other products to feature and get indexed. So you remove it from the homepage. The 3rd time the crawler comes around does not see link to the product on the homepage any more but finds that navigational category link and says, "Here is that link again, but the homepage link is no longer there." Then every time it comes back it knows there is a link on that navigational category.

OK, we know that crawlers don't talk . . . right?

In this example, the homepage link serves as a NEW LINK while the navigational category link serves as the OLDER LINK. The idea is that you are trying to communicate with the search engine. In a way you are telling it what you think is important in your website every time it comes around. Let's take that same example and go out farther in time when that product has been on your website maybe 2 years and it has been ranking #6 for the term "blue widgets" (Note: were only looking at the linking factor here - ceteris paribus). Then, you add this same product to be featured in your homepage because it's on SALE and suddenly it moves up in rankings, say #2 for example. Why? Because you are now telling the search engines (in a way) "Hey, this is one of my most important products. I want you [search engine] to see it that way." And it does, therefore it rewards you with a higher ranking.

This example was showing a single ecommerce website to simplify the differences between old links and new links. However, I believe that when search engines look at larger clusters of the webgraph and sees the same thing as it does with this example above. Using authority websites as homepages and regular sites as product pages. Our dear friend Dr. E Garcia (Orion) would call this "fractals" and yes I believe it is, because he has taught me how the web seems to be a pattern repeating itself. This is where inbound links play an important role, but even at this level the search engines should be able to distinguish old links versus new ones.
Nacho is offline   Reply With Quote
Old 03-13-2005   #8
Mike Grehan
Member
 
Join Date: Jun 2004
Posts: 116
Mike Grehan is a name known to allMike Grehan is a name known to allMike Grehan is a name known to allMike Grehan is a name known to allMike Grehan is a name known to allMike Grehan is a name known to all
Nacho,

Once again thanks for the heads up on this one. It's an interesting discussion talking about the age of links. But it really needs to have the content and context of links added to it.

Link anchor text in its usage and context can add much more about the current relevancy of web pages related to temporal analysis than pure connectivity data.

In simple terms, as I mentioned at the search algorithms session in New York, take, for instance, citation analysis. In the way that Salton referred to its usage in retrieval effectiveness, an old paper could never cite a new paper. However, with electronic documents it is possible for an old document to be updated to add and cite a new document.

I think I referred to a Microsoft research paper from a few years ago called: Effective Site Finding Using Link Anchor Information, which is very much worth a read:

http://research.microsoft.com/users/...ll_sigir01.pdf

As we're all aware since my recent conversation with Apostolos Gerasoulis at Ask Jeeves, link anchor text, in the way that it's used with connectivity data is usually more useful than the content of actual pages in terms of site finding.

And also nice to be able to use that conversation as another reason to mention that PageRank as a real ranking metric at Google is completely phoney.

I really do wish, in this industry we could just drop the whole notion of PageRank being anything other than some public relations smoke screen by Google. There are people selling PageRank and people buying it. It doesn't work, Google doesn't use it and people in this industry talking to clients about how important it is are fooling themselves and their clients. And that will lead to tears on both sides.

Ooops! Did I go off topic just a little bit there

By the way, I've had dozens of emails from people asking about some of the papers I referenced in New York. Many complaining that the presentation wasn't on the Jupiter web site. My fault, I'm afraid. They have it now and I believe it has been uploaded.

That said, here are a couple of other links I think I suggested may be useful:

PageRank HITS and a unified framework for link analysis.
http://www.siam.org/meetings/sdm03/p...s/sdm03_24.pdf

Ranking search results by reranking the results based on local inter-connectivity. (Google Patent)
http://patft.uspto.gov/netacgi/nph-P...S=PN/6,526,440

I'd love to add more, but I'm finishing another revised chapter to my book and it's nearly time for Coronation Street...

What? Yes, Coronation Street...

When I DO get back home, I never miss it
Mike Grehan is offline   Reply With Quote
Old 03-13-2005   #9
St0n3y
The man who thinks he knows something does not yet know as he ought to know.
 
Join Date: Jun 2004
Location: Here. Right HERE.
Posts: 621
St0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to all
I've never given any thought to the question of old vs. new links. Like NFFC I try to look beyond what search engines are doing and look toward what the can do and might do. This "dreaming" has given us a huge advantage in the link arena because we were going after quality and relevant links long before most others and while even still link "quality" is still being debated as to whether its a factor or not in the algorithms. (obviously, I believe it is.)

As to old vs. new, currently I don't think the search engines do weigh that, though I have no foundation or "reason" for believing that. However, being the dreamer, I think it is very possible for search engines to begin to weigh it as they continue looking for more ways to improve results.

I would think that weighing old links could be quite significant, especially when looking at external incoming links. This could eliminate "link buying" for PageRank purposes as a practice because such links are generally good for only a period of time. Any such advertisements could more easily be devalued because each time a new ad is placed its a "new" link. Heck, the engines could simply follow links but make them age a set period of time before the link is weighed into the algorithm.

I don't believe its happening, but I think that it could. If it did, we would not be effected much, nor would we have to change how we SEO in any way.
St0n3y is offline   Reply With Quote
Old 03-13-2005   #10
AussieWebmaster
Forums Editor, SearchEngineWatch
 
AussieWebmaster's Avatar
 
Join Date: Jun 2004
Location: NYC
Posts: 8,153
AussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant futureAussieWebmaster has a brilliant future
An extraordinary thread... and one I hope we can develop similiar to a few started by orion. This is an informative area that many should read to start and understand the nature of the web in less numeric terms.
Your post Nacho is one of the best I have read that simplifies yet explains how the web is referenced by search engines.

I agree that PR is a huge public relations smokescreen by Google. But its one worth is to assess the sites Google itself sees as hubs or authority sites - or at least that is how I use it and it seems to work.
The randonmess or at least the appearance of it in their listing of IBLs also seems to contradict the PR importance theory.

The carbon dating analysis (or should that be silicone dating) of web sites also gets support from the recent 'sandbox' phenomenom - thinking about that has anyone done a test of the sandbox effect on sites directly submitted to Google versus those with links placed on well spidered sites.
If dating is an important element then there also needs to be resarch done on what determines the date - has Google bought into the registrar game to aid with dating (beyond all other possible information it will garner through that)? Was the once important DMOZ part of that equation?

I have my light reading mapped out for this week (thanks for the new links - and the need to reread some others from a different perspective).
AussieWebmaster is offline   Reply With Quote
Old 03-13-2005   #11
randfish
Member
 
Join Date: Sep 2004
Location: Seattle, WA
Posts: 436
randfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to all
There are two ways I see the temporal link analysis being useful:

#1 - 15 Minutes of Fame
The time and frequency of links going to a particular site could be used to create a "15 minutes of fame" list internally or for the public by the search engines. I think this could be a big hit, as it would let web searchers and users know which particular news items, sites or pages are being referenced frequently right now. The data could be checked against spam filters to see if the site was gaining popularity out of content value or artificial link popularity inflation. You could use a training mechanism and point the algos to sites that gained link pop. for real vs. those where it was "bought" and thus improve the quality of results.

#2 - Do you still like that site or are you just lazy?
The older sites and older links could be examined against the frequency of their content and link updates to see if the links therein were continued "votes of confidence" for the long term, or just a forgotten, neglected page. This could help search engines to evaluate the quality of links even further and give insight into what the site owner's intentions were with that link.

There's probably several others too, but these were the ones that came to mind. Even just these two could have a big impact on SEO.
randfish is offline   Reply With Quote
Old 03-13-2005   #12
neuron
Member
 
Join Date: Dec 2004
Posts: 24
neuron is a jewel in the roughneuron is a jewel in the roughneuron is a jewel in the roughneuron is a jewel in the rough
Quote:
Originally Posted by St0n3y
I would think that weighing old links could be quite significant, especially when looking at external incoming links. This could eliminate "link buying" for PageRank purposes as a practice because such links are generally good for only a period of time. Any such advertisements could more easily be devalued because each time a new ad is placed its a "new" link. Heck, the engines could simply follow links but make them age a set period of time before the link is weighed into the algorithm.
The offset to this is that it would dramatically interfere with the acquisition of fresh content. Google News would disappear perhaps. When something new and important happens details get posted somewhere on the web immediately. If Nike suddenly announced they were going to start manufacturing automobiles, Air Cars, and put up a site to auction them off as they came off the production line, they would get thousands of links overnight. Would google put themselves in a position to miss out on that?

A link is a link is a link is a link.
neuron is offline   Reply With Quote
Old 03-13-2005   #13
Nacho
 
Nacho's Avatar
 
Join Date: Jun 2004
Location: La Jolla, CA
Posts: 1,382
Nacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to beholdNacho is a splendid one to behold
Quote:
Originally Posted by randfish
#1 - 15 Minutes of Fame
That woud be one VERY fast crawler. Maybe AOL if it comes out with their own search engine, they will call their crawler "Road Runner".
Quote:
Originally Posted by randfish
#2 - Do you still like that site or are you just lazy?
I'm sure anyone with a link in Tim Burners-Lee homepage would be worth A LOT no matter if it has been there since 1989 and the page hasn't been updated much.
Nacho is offline   Reply With Quote
Old 03-13-2005   #14
St0n3y
The man who thinks he knows something does not yet know as he ought to know.
 
Join Date: Jun 2004
Location: Here. Right HERE.
Posts: 621
St0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to all
Quote:
The offset to this is that it would dramatically interfere with the acquisition of fresh content. Google News would disappear perhaps. When something new and important happens details get posted somewhere on the web immediately. If Nike suddenly announced they were going to start manufacturing automobiles, Air Cars, and put up a site to auction them off as they came off the production line, they would get thousands of links overnight. Would google put themselves in a position to miss out on that?
Good point. Pros and cons all the way around.
St0n3y is offline   Reply With Quote
Old 03-13-2005   #15
lots0
 
Posts: n/a
Quote:
Originally Posted by neuron
A link is a link is a link is a link.
I agree.

I maybe wrong, but I believe that any system that would be able to categorize and record links on a page by chronological age would have to be part of the core program (like PageRank) and there is no evidence (that I have ever seen) that this is or ever was the case.



Quote:
Originally Posted by NFFC
I think SEO is an art not a science, the good bits of it anyway.
Science and art are not exclusive. Were not all the greatest scientists also great artists?
(Not compairing SEO to Michelangelo Buonarroti or anything like that )

Last edited by lots0 : 03-13-2005 at 10:25 PM. Reason: can't type
  Reply With Quote
Old 03-14-2005   #16
rustybrick
 
rustybrick's Avatar
 
Join Date: Jun 2004
Location: New York, USA
Posts: 2,810
rustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud of
Would it make sense to break out links based on where they are coming from?

Hub links are simply solid, and the older they are, the moe value they should have (IMO).

Authority links are great as well, but newer links might be considered stronger then links that are 3 months old, but links that are 2 years old might be just as strong as the new links. Why? Well, new links at an authority should cause a buzz, imo. 3 month old links might not be relevant anymore. 2+ year links should be golden, because authorities, become that way from being up to date and should not contain many broken links (although, you and I know they do).

Every other link? a link is a link.

Not sure if this adds at all.
rustybrick is offline   Reply With Quote
Old 03-14-2005   #17
lots0
 
Posts: n/a
Quote:
Would it make sense to break out links based on where they are coming from?
Sure it would, thats is why there are "bad neighborhoods", like the searchking network.

But that has nothing to with the age of a link.


Quote:
Hub links are simply solid, and the older they are, the moe value they should have (IMO).
I agree, they should, but I have yet to see any evidence that they are.
  Reply With Quote
Old 03-14-2005   #18
rogerd
Member
 
Join Date: Jun 2004
Posts: 109
rogerd is a jewel in the roughrogerd is a jewel in the roughrogerd is a jewel in the roughrogerd is a jewel in the rough
Interesting topic. As NFFC notes, one can make a case for new links AND old links. Maybe middle-aged links can serve a special purpose, too.

This strikes me as one of those variables where the SE wonks twiddle the dials and see what happens. Logically, weighting recent links heavily might favor spammy sites, but not doing so might miss great new content, too. One could philosophize about this forever, but different strategies would have to be run against actual sites to see if the cream (or at least a reasonably good dairy product) rises to the top.
rogerd is offline   Reply With Quote
Old 03-14-2005   #19
2much
Member
 
Join Date: Feb 2005
Location: Los Angeles
Posts: 24
2much will become famous soon enough2much will become famous soon enough
I've always worked using a combination of both.

I have a base number of links that remain for the site permanently. Then, we continusouly add new links using different methods, but alwasy new links are added, every month.

Then I take a subset of sites and change the links. So, every month, the links get switched and the target sites end up with both old and new links. It keeps the structure both old and "fresh". Also, in switching up the architectrue, we never drop in # of inbound links, but always grow.
2much is offline   Reply With Quote
Old 03-15-2005   #20
St0n3y
The man who thinks he knows something does not yet know as he ought to know.
 
Join Date: Jun 2004
Location: Here. Right HERE.
Posts: 621
St0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to allSt0n3y is a name known to all
our theory is that new links eventually become old links so we've got the best of both (if there is such a thing).
St0n3y is offline   Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off