110 replies to this topic

[Scottie]

Welcome customs!

The general consensus around here is that those are great links to have, just not because they have a .gov or a .edu ending, but because they are generally respected sites.

The debate isn't over whether or not you want links from those sites... you do! No one's going to say turn 'em down. But it's not the domain extension that gives it any magical properties.

[pageoneresults]

It's pretty amazing to see how this topic has progressed. I've read through the entire thing twice just to make sure I wasn't missing anything. All these below statements can only mean one thing, .gov and .edu links out of the box have more value. They are Trusted TLDs in the general sense.

Weary, Yahoo links are helpful because they’re high PageRank, but that’s the only reason; there’s no special "Yahoo boost" or edu-boost or gov-boost. Those links just tend to be higher quality.

There's no boost just for having those TLDs, but the sites with them are generally high quality and authoritative.

They're not high-quality and authoritative because of their TLD, but their TLD, because it's only given to certain types of sites, is often on sites that are authoritative.

Of course, this comes with the massive post clause that all links help, and .edu and .gov links are likely, on the whole, the be amongst the best types of links without needing a specific boost for simply being .gov and .edu, as these sorts of sites are likely to be well linked to naturally.

One could argue that a .edu is more likely to be an authority, simply because the restrictions in the creation of the domain. I believe this to be true. This, however, is a correlation, not causality. By its nature, a .edu site is more likely to have resources to develop authoritative content, will be hosted on a quality network, will be persistent for a number of years, have some name recognition out of the gate, etc. All of these characteristics will help a site in aquiring links and becoming an authority.

The general consensus around here is that those are great links to have, just not because they have a .gov or a .edu ending, but because they are generally respected sites.

If the above is not an indication that .edu and .gov links are of a higher value, I'm not sure what else I could provide that would be more convincing than those responses from everyone participating in this topic.

[sim]

While I've tended to lurk on these forums, I decided to finally register so I could add my 2c worth to this particular thread. No evidence, just some food for thought based on personal experience.

I am employed as a web developer at a University, a position I have been in for 5 years, and while all Universities are different and have different priorities regarding the importance of their websites, the one I work for uses a central content management system (which most here seem to use now). While there is a central web office to maintain the server with the CMS, no central web development is offered, and a couple of different multimedia groups like mine exist on campus within different faculties to provide assistance as needed on a user pays system.

The departments / schools / faculties all have very tight budgets, and generally don't want to pay to subcontract my skills when they can get their admin staff who have no web development skills (and not necessarily great computer skills) to do the work for no extra cost (there are exceptions of course, otherwise I wouldn't have a job still!). As a result, the general quality of the content of the sites (not to mention the navigation etc) is terrible. Consequently, most of their rankings are likewise terrible, so it is unlikely that anyone outside the University would stumble across the information, as it is difficult enough for those of us within the University to find any information on the sites...

Most of the department / research group sites within my University have no links within the content, but a separate "Links" page with a collection of links related to their department / research group's topic, usually not categorised at all, but just listed in one long list, very much like how many web sites were constructed in the 90s. Once the links are added, it would be unusual for the administrative staff to go through and check that the links still work, so whether a page with mostly broken links would add any credibility to the sites that they link to seems doubtful to me (and how useful is it to a human who stumbles across the site and then has to scroll through screenfulls of links that aren't in any sort of order?).

So while some University sites might be good authorities and provide useful content and good links, it wouldn't be true to generalise and say that all University sites are good authorities, including those of legitimate departments / groups.

Also a comment about .gov links... In Australia, government departments change their name on a frequent basis, as they are merged, split, privatised etc, and each change sees them throw out their old domain name and aquire a new one - and they don't bother to keep the old domain with a redirect to the new, which is surprising (another cost cut?). So here, .gov.au domains and URLs change and disappear all the time - more often than .coms! I doubt that would help any outbound links much?

[pageoneresults]

Darn, I thought I had it sold and along comes sim, out of lurking mode to rain on the parade. Thanks!

[sweepthelegnate]

Darn, I thought I had it sold and along comes sim, out of lurking mode to rain on the parade. Thanks!

you had me... (not that you ever didn't though.) but now sim makes some excellent observations and the quality not necessarily reaching across the whole spectrum.

[projectphp]

The general consensus around here is that those are great links to have, just not because they have a .gov or a .edu ending, but because they are generally respected sites.

If the above is not an indication that .edu and .gov links are of a higher value, I'm not sure what else I could provide that would be more convincing than those responses from everyone participating in this topic.

Welcome to the wonderful world of semantics. I hope you find the meaning you are looking for (semantic jokes KILL at Uni).

So, Edward is now in complete agreement with everyone else; most likely, although not always, certain restricted TLDs, gov, mil and edu as examples, will tend to provide better link benefit, on average, than any random selection of non restrictive domains.

However, and here is the point that Edward seems to be missing, this is not the result specifically of being such a TLD, but, as Jill mentioned on like page two, a result of such sites being exactly what SE algorithms are designed to find and promote: well connected authorities with many links in from all over the web. In other words, if stanford moved to stanford.com, or rather had always been stanford.com, its links would have exactly the same weight as they now do. The .edu bit offers nothing special over any other TLD.

Can everyone agree on that, or do we need 4 more pages, and about seven brazillion more semantic argument? Cause I am so up for that you have no idea! The philosophy course I am taking is having a week break, and I needs me some semantic argument fix. having spent half a semester arguing the existence of God, something like this should be a piece of cake

[pageoneresults]

So, Edward is now in complete agreement with everyone else; most likely, although not always, certain restricted TLDs, gov, mil and edu as examples, will tend to provide better link benefit, on average, than any random selection of non restrictive domains.

However, and here is the point that Edward seems to be missing, this is not the result specifically of being such a TLD.

Arrrggghhh, I'm so confused! What exactly did you just say there? Is that somewhat contradictory?

[qwerty]

I'd respond, but I defined a term for someone last week and it had to be pointed out to me that they'd asked what the word meant as a joke, so as I seem to have lost my eye for sarcasm, I'm keeping my trap shut

But I will say welcome to the new members and thank you in particular to sim for a very useful perspective on the discussion.

[projectphp]

It is not a contradiction at all!

Cause and effect. Is it a good link just because it is .edu, or because of the same reasons any other link is good, i.e. that the site is well linked to?

The examples given, of student, faculty and part-time faculty, all point to the fact that .edu pages are not liklely to always, 100% of the time, be a great, trust worthy authority page.

Which leads to the question is the algo picking out good pages, irrespective of TLD, or is TLD alone a factor? For this question, there is absolutely no evidence either way, and by Occam's Razor:

entia non sunt multiplicanda praeter necessitatem, (which translates to:entities should not be multiplied beyond necessity).

the simplest answer is that .edu, .gov and .mil are treated just like every other TLD, and are good links for the same reason every other site is or is not.

Make sense?

[pageoneresults]

Make sense?

It does. But, it doesn't convince me that they are not Trusted TLDs and have more value with all other factors being equal.

Cause and effect.

Cause = The producer of an effect, result, or consequence.

Effect = Something brought about by a cause or agent; a result.

The Cause? A Trusted TLD.

The Effect? Generally of Higher Value in the overall picture.

All TLDs are going to have their "neighborhoods". Obviously student pages that are not well maintained and/or don't meet the basic guidelines are not going to perform as well. If we go back to some of my initial replies I asked that we look at the core TLD and static pages (I should have defined those that reside outside of the student areas). But, that doesn't necessarily mean that all student pages are not of value.

Just take a look at the latest restrictions on the .edu TLD. It becomes more trusted each time they tighten their guidelines. And yes, semantics have an important role in all of this. There ya go, I said it.

Can everyone agree on that, or do we need 4 more pages, and about seven brazillion more semantic argument?

[qwerty]

Let me muddy the waters a bit here...

A client of mine just built and published a site behind my back in order to grab traffic for a couple of very important keywords. Had I known, I'd have told them not to do it, but that's not the point.

The site is coming up at #1 on Google for a number of geo-targeted versions of those keywords, and he asked me why it was coming up over the site I worked on. I was going to say its a new site and its very common for new sites to pop into the top of the SERPs, but to vanish very quickly.

Then I noticed that the SERP is linking to the site's IP address rather than its domain name, and that only the IP address version of the site is displaying any green on the ol' PR toolbar.

So I checked backlinks on both versions of the site on G, Y an M. G reported nothing, but the other two linked to a deep page that linked out to many other sites about Chinese art, and it was linking to a gallery in Beijing using what must have once been its IP address... which is now my client's IP address.

And guess what -- that deep page just happens to be on a .edu site. So there you have it. Absolute proof


(That "absolute proof" thing, for those who are as blind to such things as I apparently am, is sarcasm, by the way.)

[projectphp]

But Edward, why does an SE need the TLD to decide something is trusted? What makes you think that the TLD matters more than anything else, or even at all?

You seem to want to change the argument. The question was simple:

Are there any evidence that having inbound .edu and .gov links are "better" links and counts more than other links ?

And the answer, still, four pages later, is no, there is not any evidence that a link from crap-y-u is any better than a link from a page with a similar pagerank et al on another TLD. None whatsoever.

I remember many an educational institute falling afoul of the authorities, including Dr John "men are from mars, my PhD is too" Gray's. To say that an algorithm that maps all the links on the web would use a TLD ahead of the brazillians of links it knows about seems to me to be, well, a flawed argument.

Put practically, a link from that great bastion of fit news printers, The New York Times, is a better link than one from the vast majority of edu sites, because TLD isn't the only or most important factor. IMHO, the importance of most specific TLD sites can be explained without assumming it is the TLD that does anything at all.

But as I said, there is no evidence that Google do or do not give certain TLDs a helping hand, and lack of evidence should mean that the least complicated idea (TLD doesn't matter) wins, but that we keep an open mind because you never know!

Now, if anyone finds any evidence, please let me know, because I am willing to listen...

[champa]

I agree with Jills view. Sites link Yahoo Directory, CNN, BBC, DMOZ, ZDNet, National Geographic, PBS.org are white listed in search engines. Links from those white listed sites carry more weight than other normal domians. Otherwise weight of link depens on PR of the site not on the extension (gov, edu stc) of the site.

[Dodito]

First of all: my background is not SEO at all. We are building a portal and therefore SEO is of interest to me.

My own background (years ago) is mathematical modelling in physics.... and so I feel I could make a more general point here that most (not all) seem to be missing.

Unless you have the equivalent in mathematical/statistical calculational power to perform tests on their algorithms that are statistically sound with a decent margin of error.. you're absolutely wasting your time.

A few dozen examples where a hypothesis "works" do not constitute enough evidence to proclaim a general "rule" like "keywords in url's work" ... or "links from .edu and .gov sites are better" )

Does it mean pattern recognition we call "intuition or gut feeling" and applying "common sense" aren't useful ? Ofcourse not.

But don't pretend to find the answer in the smallest detail (your margins of error by FAR surpass any precision you claim you can find with a few dozen examples) and just apply what makes most sense given what Google and Yahoo! want to achieve.

Question in case: do TLD's like .edu and .gov help ? Yes ofcourse.. why ? the ONLY answer you can give is "we don't know". It could be because of the # backlinks they receive themselves, because it's a trusted organization, because they generally have a high PR, because Google/Yahoo will be able to differentiate a professor's site from a student site or... a million other reasons.

And common sense tells you it's a mix of reasons.. maybe the weight factors even differ per area (that is what I would do to really screw up the SEO community that wants to tweek the ranking of a site based on these very generic assumptions. A ranking which in competitive keyword area depends so sensitvely on so many different factors you would never be able to figure it out for an individual parameter anyway with trial and error no matter how hard you tried.).

In other words.. sometimes you'd score wel with some general very specific rule.. and sometimes not.. (and isn't that what happens a lot of times ?) (ANd I am not talking about the more common sense issues such as "good content"; "many valuable inbound links" etc etc.. I am talking about these specific ideas such as keywords in TLD's or 5 year registration in the WHOIS or whatever..

So......back to the beginning: unless you have the mathematical prowress to do a thorough statistical analysis.. all explanations are BS. What matters is that it counts.. and so do links from many other sites. To pretend it's a science is not only fooling your client but yourself as well (if you believe your own BS !).

So.. my conclusion is: you'd probably be succesful with two groups of sites.. (but I am not an SEO guy whatsoever): small sites in niche areas.. you can gain a lot by applying common sense rules.. and the owners usually do not have the time to learn/understand everything. The other group is the super authority sites.. where you can use "overwhelming force" and apply any rule that can help somewhat.. to let them score high even in competitive keyword areas (but you could never guarantee a ranking since your level of uncertainty of what works and what doesn't is too high. All you can hope for is that overwhelming force is so big.. it kicks your site in the top 5 or something).

Just my two cents..

[Michael Martinez]

Now, if anyone finds any evidence, please let me know, because I am willing to listen...

Tch, tch! Matt's comments are evidence against such special boosts. You just arbitrarily dismissed them.