The SEMPO Website and its Page Rank Distribution

[Marcia]

Since SEMPO is a an organization formed for the purpose of representing the SEM/SEO industry, it seems like it would be interesting for those of us who have spent a bit of time around Google forums and seen all the discussions on links, navigation and PR, to take a little closer look at the SEMPO website, particularly with regard to the internal navigation structure, the outbound linking, and the PR distribution, both for the site itself and in regard to how it's impacting the outbound links from the site.

Per Yahoo! - 5,500 inbound links to the homepage
Also per Yahoo! - 6,500 inbound links to the entire site (linkdomain: )

SEMPO Homepage: PR7

In addition to the homepage links direct to the Circle Members' sites, there is a link going to this page from the Site Map, which appears to somewhat parallel the links appearing on the homepage to a degree:

Committee Members - also PR7
http://www.sempo.org/sempo-committee-members.php

SEMPO MEMBERS DIRECTORY

When clicking on that main navigation link at the top of the homepage to the Members Directory, it goes directly_to_ here:

SEMPO Members Directory - CIRCLE: PR6

http://www.sempo.org/sempo-membershi...ory-circle.php

Then, on that page are links to these

Executive Members: PR4
http://www.sempo.org/sempo-membershi...ve-members.php

Contributors: PR4
http://www.sempo.org/sempo-membershi...ntributors.php

Supporters: PR4
http://www.sempo.org/sempo-membershi...supporters.php

REGULAR MEMBER LISTINGS
Now we get to take a look at the listings for the regular members. On that main Directory page (the PR6 one) are links to other members by the alphabet, ascending and descending, and by individual alphabet letters.

If you choose the radio button for "All" this is the URL
http://www.sempo.org/sempo-membership-directory.php

Taking a look at a few by letters picked at random:

R - PR5
http://www.sempo.org/sempo-membership-directory.php
D - PR5
http://www.sempo.org/sempo-membership-directory.php
W-PR5
http://www.sempo.org/sempo-membership-directory.php

HEY!! wait just a minute there! Those all have the same URL!

So I checked the Google cache - dated August 11, 2004

D
http://66.102.7.104/search?sourceid=...-directory.php

R
http://66.102.7.104/search?sourceid=...-directory.php

W
http://66.102.7.104/search?sourceid=...-directory.php

By golly, the cached pages are all the same, too! They are not in the cache by letter at all, are they?. So for people listed on other pages, like the K or S or T pages - what's the PR of the page linking to them, and what kind of PR "vote" are they getting from the SEMPO members directory?

Anyway, who did all of that? Was it the SEO who's got a link from the bottom of every page on the site, or was the navigation orchestrated by someone else?

There is definitely more to see, but this seemed like a good enough place to start a bit of a site analysis. I have to admit that I'm feeling a little bewildered at this point about the navigation and PR structure and the rationale behind it; maybe I'm missing something here. If so, hopefully someone will help me along by explaining how all of this is supposed to be working.

[polarmate]

Simple. Spiders don't follow forms and don't follow javascripts. The alpha listings are generated by a script that submits the letter to the form.

<script type='text/javascript'>
    <!--
    function setSelectBy (x) {
        document.getElementById("selectby").value = x;
        document.showMembers.submit();
    }
    // -->
    </script>

where x is A,B,C,...

It's pretty well thought of.

The web site design used to say webmaster Tom Dahm when th site was first launched. He is also the domain owner for bridgepose.com.

[Marcia]

polarmate:

Simple. Spiders don't follow forms and don't follow javascripts. The alpha listings are generated by a script that submits the letter to the form.

Ohhhh, I see! Thank you.

So then, technically all of the total membership altogether all have one link to their sites, only from that one PR5 page. Not much PR to that, is there? Where are those Page Rank maths gurus when we need them?

Moving along...

Additional SEM Resources - PR6
The links on the page are worth a look.
http://www.sempo.org/additional-resources.php

Then, linked to from that Additional Resources page:

Search Engine Marketing Forums and Blogs: PR0
<meta name="robots" content="noindex, nofollow">

SEM Information Resources: PR0
<meta name="robots" content="noindex, nofollow">

SEM Tools & Software: PR0
<meta name="robots" content="noindex, nofollow">

There are a couple that *seem* to have PR6, but those are anchored links on that same page, so it isn't all that obvious to catch which are carrying PR and see that some have been robots.txt excluded.

The PR distribution does appear to follow a pattern, and looking through most of the content, i.e. case studies, articles, etc., as well as the links to the members themselves, a pattern does seem to emerge, indicating that not only is a good portion of PR selectively siphoned off but also that the meting out of Page Rank throughout the site was done by design.

More of a picture emerges when looking at the site map. With all the different issues being brought up, it just seemed like it was time to start to take a closer look at the site itself.

[robwatts]

Looking at those google member list caches I'd be very upset if I were this member

table cellpadding=

no link at all for that poor bugger

Seriously though, It would be interesting to hear the reasoning behind why they used noindex, nofollow on those other pages. A robots.txt file could have achieved the same of course, but why, why exclude those pages, for what purpose or benefit? I'd understand if they were secret resources, or maybe its part of some cunning experiment??

The sites webmaster has a pr0 , it is a new website though as it was only registered on March 14, 2004. Odd that SEMpo have allowed a link to their homepage on each and every page. Yeah , yeah I know standard "designed and hosted by.." practice and all that, but from an ordinary members POV, they might argue that its unfair that the webmaster is receiving double benefit, paid for the job, and rewarded with a link on each page too, same applies for the hosts too...unless of course the job was done buckshee, in which case its all academic...or is it? In light of all that NPO stuff maybe it isn't.

The more things like this are scrutinised the uglier it all seems, especially with no response from any official channels on many of the issues raised elsewhere.

Sad really.

Oh yes. It's all done by design and certainly on purpose. It's really just another way that Circle members are benefited the mostest.

I find the fact they are using a noindex, nofollow tag on all the resources pages quite disturbing and blatant and on purpose Crap.

Of course, this goes hand in hand with things:

BTW, I don't believe for an instance this is Danny's doing at all.

On the left hand navigation bar is this:

"Forum Roll"

Listed are these forums:

Cre8aSite
HighRankings
JimWorld
SEO Chat
Search Guild
Webmaster World

One forum which is the 3rd oldest forum on that list is missing. Only WMW and JimWorld are older. It's especially disturbing when the forum owner is directly linking to the SEW home page on the forum owner's new website. Of course, this must be an accident.

[dannysullivan]

Doug, if that's a request to be listed in the forum roll, then PM me with a proper request or send me an email. It's off topic for this thread.

[Marcia]

>>Looking at those google member list caches I'd be very upset if I were this member

Apparently not the whole page is being crawled (going by the live page linked to from the cache) because of the error or it would show in the cache - else there would be what, 300 links on that one PR5 members page? At least half aren't showing in the text-only cache. Looks like NO links to those sites to me.

Either that or they just don't index that many links on a page - Google themselves DO say to limit pages to 100 links.

Then, what really gets me is that the individual alphabetic pages LOOK like they've got a PR5 unless you happen to catch the URL being the same - and they are deliberately being excluded by using JS.

For Gods's sake, one regular members' membership fee could pay for a readily available link script that puts out static, crawlable pages. This is all EXACTLY the type of things we've always advised "beginners" asking how to check out the value of links what to watch out for.

And what do we figure when we see JS links and <noindex,nofollow> on link pages? We figure it's PR hoarding, and people in forums are generally advised to watch out for that also.

Don't people check these things out? That's what we're SUPPOSED to be doing as SEOs - it's part of our JOB!

Not really though Danny. The resource page has 'all' forums listed but with a noindex, nofollow as well. I'm just pointing out other stuff that goes along with Jupiter Media.

No need to PM you for something that is obvious Danny. Like I said, I know it's not your doing.

Yes Marcia, it's PR hoarding at best. That is very obvious. I guess to answer your question; there are soooo many things wrong with SEMPO that there are just too many things to research. One would assume (wrongly) that a group representing the industry with "Professional" in their business name would conduct themselves as a Professional from the building of the website. Now we know it's not the case and never has been the case with sempo.

[Marcia]

Doug, I don't believe for one single second it's got anything whatsoever to do with Jupitermedia. I don't know who did it, or where the ultimate responsibility for this lies, but I think it's pretty obvious that no one checked it out.

[dannysullivan]

It's all done by design and certainly on purpose. It's really just another way that Circle members are benefited the mostest.

These are the people who also paid the most, of course. Whether there should be different levels like this is of course being discussed in another thread here: SEMPO & Membership Benefits.

In terms of getting a link on a high PR page, it's not that helpful if the link is merely your company name. That definitely will help ensure these companies improve the odds of being found for their own names -- but chances are, that's hardly a benefit they need to join SEMPO for.

Instead, you'd rather have the additional listings allowed on the actual member directory page. There, Circle members are allowed to have some additional and descriptive links. That lets them pick up a link and point at a page saying things like "search engine marketing services" and so on. Now that's a real benefit -- high PR and link context/anchor text. But I'd say it's still far from turning the SEMPO web site into a link farm, which some have accused it of being.

It would be interesting to hear the reasoning behind why they used noindex, nofollow on those other pages. A robots.txt file could have achieved the same of course, but why, why exclude those pages, for what purpose or benefit? I'd understand if they were secret resources, or maybe its part of some cunning experiment??

I agree -- there seems to be no good reason no to let these pages get indexed.

but from an ordinary members POV, they might argue that its unfair that the webmaster is receiving double benefit, paid for the job, and rewarded with a link on each page too, same applies for the hosts too...unless of course the job was done buckshee, in which case its all academic...or is it?

Yes, it may very well be that the web designer and hosting company weren't paid but instead got a credit. I honestly don't know. I do know that there's been strong evidence that search engines may not weigh as heavily a link they see in the same way on every page. And in both these cases, the links again contain the company name, minimal value if you were aiming for some particular search engine benefit.

[dannysullivan]

Not really though Danny. The resource page has 'all' forums listed but with a noindex, nofollow as well. I'm just pointing out other stuff that goes along with Jupiter Media.

SEMPO is not owned or operated by Jupitermedia, Doug. They don't control what SEMPO puts on its resource pages over there.

If you have an issue with wanting to be on the Forum Roll on THIS site, which is run by Jupitermedia, then PM me or send me an email. It's not on topic for this thread.

Don't people check these things out? That's what we're SUPPOSED to be doing as SEOs - it's part of our JOB!

People ask me if they should buy links. I say sure, if you feel the traffic from the link itself will justify the fee. I say don't buy a link because you think there's any search engine specific value you think you may gain. Sure, that may come -- but so many things can also complicated it.

So this is ironic. Some people might join SEMPO thinking, "Hey, I'll get a link, that will help me with search engine ranking." That wouldn't be my advice. But let's say it does. Then SEMPO gets set up for accusation of being a link farm. But then if people are on pages that don't pass along PR value, that accusation goes away -- but they also get opened up to hording PR!

But I think your point is well taken, Marcia. If you're going to have links on your site, shouldn't it generally be one way or the other. All the members should get whatever PR value that's going to get passed on, not just some. Whether accidental or on purpose, it's something that should be fixed.

hmmm Danny. A link is a link whether or not it's just your company name or a keyword phrase in the link text. Of course a link from a good PR page is going to help the "linked to" website. A little boost is given if the text is a phrase, but the benefit of the link itself is certainly a search engine benefit that cannot be ignored.

[dannysullivan]

A little boost is given if the text is a phrase, but the benefit of the link itself is certainly a search engine benefit that cannot be ignored.

"Boost" can't be divorced from the anchor text. If it was just about getting links, a search for anything would bring up Amazon tops. It doesn't. But a search for "books" does, because lots of people link to Amazon with the word books in or near the link to the site. There's a world of difference between getting a link and getting a link with anchor text that matches to what you want to be found for. You might want to check outEffectiveness of Anchor Text and Change To Link Bomb Sign Of New Link Analysis Shift?, which are recent threads that cover this.

I respectfully disagree. There is a "little" boost for the phrase in the text. That's a given. There is a BIG boost for the actual link itself no matter what the heck the link text says.

Putting it this way:

I quality link on a quality site that leads to your website is MUCH better than no link at all. I truly don't give a hoot what the link text may say. That's not as important as the link itself.

[Brad]

Don't people check these things out? That's what we're SUPPOSED to be doing as SEOs - it's part of our JOB!

Marcia, SEMPO was banking on trust, so nobody questioned it or looked closely. That trust in SEMPO is being used up (wasted) at an alarming rate, so the scrutiny of everything SEMPO and the directors do is only going to intensify.

[polarmate]

Apparently not the whole page is being crawled (going by the live page linked to from the cache) because of the error or it would show in the cache - else there would be what, 300 links on that one PR5 members page? At least half aren't showing in the text-only cache. Looks like NO links to those sites to me.

Either that or they just don't index that many links on a page - Google themselves DO say to limit pages to 100 links.

Marcia, Google only indexes the first 101k of a page. That page is over 200k. It's not by the number of links but the page size.

The other problem with that page is that those alpha links work only in IE. They don't work in Firefox 0.9, Opera 7.23, Mozilla 1.7.1 - didn't bother with any more browsers as the reason was obvious. IE does not care about case when it comes to names and id of objects and elements. Other compliant browsers do. The object is given the id 'selectBy' but is referenced in the javascript as 'selectby' which is a no-no if you want to ensure cross-browser compatibility.

[Marcia]

Ah, I didn't think about the page size - I've had to work with pages of close to 300k - but it was all graphics and very skimpy on text, around 20k and heavy with code. Then there are the users who surf with JS disabled altogether, aside from the browser issues you mentioned. I hadn't checked others, only IE.

I believe it was NFFC who recently posted that when someone is digging themselves into a hole the important thing is to stop digging. Unfortunately, it looks like the site is a big shovel.