homepage Welcome to WebmasterWorld Guest from 54.167.182.201
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google SEO News and Discussion
Forum Library, Charter, Moderators: Robert Charlton & aakk9999 & brotherhood of lan & goodroi

Google SEO News and Discussion Forum

This 56 message thread spans 2 pages: < < 56 ( 1 [2]     
Common Google SEO Myths
grant

10+ Year Member



 
Msg#: 3501195 posted 1:07 am on Nov 10, 2007 (gmt 0)

I'd like to point out what topics I hear all the time that I believe are myths, and explain why.

1. Myth: Time on site matters
Some of the most valuable sites to me are the ones I stay on the LEAST amount of time, because I got my information FAST.

2. Myth: Bounce Rates affect my ranking
When I can get the information I am looking for on the page G finds most relevant, voila! No need to go any further! Awesome!

3. Myth: Natural link text is important
The most natural link is either “click here” (Adobe ranks #1 for Acrobat reader) or the name of the company. “Varied” link text is actually unnatural.

4. Myth: Link acquisition rates shouldn't be too fast
News worthy items generate a ton of links quickly. Example: the host of a future Olympics is announced. A site (Olympic site) that is new, had no links, suddenly has a ton overnight.

But one could argue these are from “authoritative” sites and won't raise a flag. Ok, think of all the esoteric niches that have the equivalent of a major announcement happen.

By ignoring or even penalizing these links, Google does the opposite of serving good results. I think Google LIKES seeing sites that get links fast, it is a clue that something is "hot".

5. Myth: Links from topically relevant sites are critical
Search for “financial services” and Adobe appears #5. Why? Because their homepage (PR 10) links to a page with the words “financial services”. That’s it. And they are beating every single financial services company for that term except Primerica. And don’t you think AMEX, VISA, and all the banks get links from “topically” relevant sites? Face it, Google doesn't know topical relevancy, it is too much of a wishy washy thing to comprehend for a computer. It understands PageRank.

6. Myth: PR doesn’t matter
See #5 above. Adobe is a great case study on how PR works.

 

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3501195 posted 4:30 pm on Nov 13, 2007 (gmt 0)

Any idea what <!--googleoff: index--> is referring to in the source?

googleoff: and googleon are special Google Search Applicance tags [code.google.com] that work like Adsense section targeting does for the Adsense bot. Those comments have no effect on regular google.com indexing. For the Google Search Appliance or Mini, they can take four different attributes, index, anchor, snippet, and all.

martinibuster

WebmasterWorld Administrator martinibuster us a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month



 
Msg#: 3501195 posted 6:00 pm on Nov 13, 2007 (gmt 0)

And has the young lady been seen since?

Marissa Mayer
Vice President, Search Products & User Experience

glengara

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3501195 posted 6:06 pm on Nov 13, 2007 (gmt 0)

Disinformation then :-)

MrStitch

5+ Year Member



 
Msg#: 3501195 posted 10:25 pm on Nov 13, 2007 (gmt 0)

It's interesting to hear peoples thoughts on 'bounce rate', as I remember ranting about this a year ago or so, when I was complaining that a site ranked #1 for it's prime term, yet the site was so god-awful ugly that I couldn't imagine anyone spending more than 2 seconds there.

Just checked... Yes, that pile of crap is still there.

So it sounds like the bounce rate is something that G would like to measure and take into consideration, however they're having trouble doing so.

How 'bout this as a suggestion: Disregard a bounce rate entirely when a user views then leaves (obvious reasons why were discussed earlier). However, you could weigh the following process -

1) User searches for a specific term (lets say Video Cards)
2) User clicks on a link in the G serp.
3) User spends some time on the page, and then clicks further into the site, by many pages.

In this case, we would disregard the serp result, however... it does tell us that the user liked what he saw and moved further. In other words, the site in question may be a quality site to serve the users for it's 'more general' term (such as 'computer parts').

To me, this makes total sense, but G would have to solve a few 'gaming' problems, such as webmasters buying traffic, etc. Some measurements could be put in place, such as how much readable content is on the page vs. average time for a person to read something... er... whatever.

On the other hand, G wouldn't have to measure ALL the traffic coming in. Maybe just a percentage of random IP's, also weighed against relevant sites in the same search result. Would have to remember that lower search positions will not get as much traffic.

G could take into account 'traffic pattern' as well. A group of users (a sudden 'bubble' on one day), clicking the same links through the site could be interpreted as paid traffic. Or traffic from one specific site, combined with time on site.

Should also take note that G doesn't necessarily have to count 'all' traffic, in terms of bonuses or negatives applied to the site. It could just group some random numbers and say "Is this group worth analyzing?" Yes/No - Move Forward/Toss in trash.

G could grab snap shots of data sets on random days through out the week too. So, someone wanting to game the system would have to buy traffic, every single day, and a considerable amount from various sources, keeping it all 'copesetic' to their current/past traffic values..... i.e. - forcing the webmaster to spend a whole lot of cash and time to acquire traffic.

Oh, and one more thing (getting winded here), G would have to arbitrarily ignore traffic from the same IP that logged into analytics. Right now, I believe you have to physically tell it to do this.

Think I covered everything. Lets hear some ways you could game this type of system.

MrStitch

5+ Year Member



 
Msg#: 3501195 posted 10:58 pm on Nov 13, 2007 (gmt 0)

Done with the analytics rant... now on to something else! ;)

So many times have I heard the pro's say 'PR is worthless', don't bother with the green pixels. etc.

The case breaker for them is "Look at this serp. That PR 0 site is ranking #1. So the green bar obviously isn't worth anything"

Yet at the same time, if you check their backlinks, they may have some real good ones, from high PR sites. Combine that with the fact that TBPR is nothing more than a 'snapshot' from god-knows-when, and it becomes plainly obvious that the REAL pr for this PR 0 site, is actually much higher.... we just don't know exactly 'how much' higher because of the old data in the little bar.

In effect, seo guru's shoot themselves in the foot with their own example, because they tell you to ignore the green bar because it's old data.

On the discussion of buying links from high pr sites, and whether or not PR is the factor..... I haven't had it completely spelled out before me, but speculation from several different conversations, from several different guru's (if you wanna call 'em that), over the past couple of years, has pointed to the fact that these same guru's buy links for the exact same purpose, under very similar circumstances..... just done in a different fashion than looking at the green bar (though, the green bar is almost a perfect reflection of what their doing anyways, so it's a wash)

In the end, you could honestly say "Ignore the green bar, cause it's not the whole story".

However, it IS very much so a part of it, and probably bigger than they care to admit. Why not admit it?

1) SEO pro's don't wanna tell you that, cause you'll probably end up competing with them at some point in time, if not.. right now. Second, and most important, they aren't going to make money by telling you how it's going to get done. Heck, one VERY highly recognizable seo (and everyone here probably knows him), charges a hellashish amount for their services. The same person is going to use that money to buy links. What's left over, is all theirs.... and I &^#! you not... thats ALL they do.

2) G isn't going to tell you PR is a factor, cause they got their serps to look after. Bad serps = loss of users, then loss of money.

In a nutshell, the people who know (or have years of experience to make a darn good guess), WANT you to ignore this cause they BOTH got something to loose.

Now, granted.... I'm a still a newb at this. But with every serp I've seen, every document I've read, every conversation with 'larger' people, has got me thinking that it's all just smoke and mirrors to help blow people off the path.

I would like to believe in the guru's, as I'd like to have a good quality source to look to when I have serious questions. Yet, I feel like it all really boils down to a 'every man for himself' kind of mentality.

BradleyT

10+ Year Member



 
Msg#: 3501195 posted 8:06 pm on Nov 14, 2007 (gmt 0)

The problem with bounce rate and measuring time is this -

If I search for a simple to answer query like, "Who was the 25th president of the u.s.?" I can probably get the answer without clicking any results.
coincidentally I picked that president number at random and the #1 result is for the wrong president. LOL?

There are many other queries where a click is needed however the result can be found within seconds on the clicked page. When I search for something, click result #1, view the page for 8 seconds and go back to google and search for something else - did I get the result I was looking for in those 8 seconds or did I give up on that query?

MrStitch

5+ Year Member



 
Msg#: 3501195 posted 10:23 pm on Nov 14, 2007 (gmt 0)

Thats exactly why I mentioned that G should ignore single entry-exit users, cause they may have found exactly what they were looking for.

If you see a user go beyond that page in the site, it could show that the site 'in general' has something to offer, etc. Therefore, there should be no 'bonus' (for lack of a better term) applied towards that one single page, but possibly to the entire site for whatever it's more broader market niche is.

Keep in mind, I'm not talking about just one single click going deeper/elsewhere into the site. It could be measured by multiple clicks, time on page, random IP's from random days, all weighed against current and previous site usage, and AGAIN weighed against past site structure/design.

In theory, this should be real easy to test out. Just have the math geeks attack the algo, setup on a different data center, and choose a few keywords, both single term phrase and double term phrase. See how the results pan out, and vwala.

Keep in mind that long tail search queries probably won't give you accurate results.

Funny that the query you made turned up the wrong answer tho... hahahha

youfoundjake

WebmasterWorld Senior Member 5+ Year Member



 
Msg#: 3501195 posted 10:39 pm on Nov 14, 2007 (gmt 0)

I actually looked at the result, and it explains why it was wrong... interesting, don't know if i agree with the logic, but...
You may have noticed that the numbering of the Presidents on this site differs slightly from other sources. The reason is because of the two non-consecutive terms that Grover Cleveland served in the late 19th century. When Cleveland began his first term in office he was the 22nd person to hold the office. When Cleveland was defeated by Benjamin Harrison in the next election, Harrison became the 23rd President.

When Cleveland defeated Harrison in the next election and took office for his second term, there still had only been 23 individuals since George Washington to hold the office of President, Cleveland being the 22nd. However, because of Cleveland's non-consecutive terms some sources call him the 22nd and 24th President. This causes the numbering on site.net to differ by one for every President after Cleveland. However, even though George Washington served two terms no one calls him the 1st and 2nd President. The fact that Cleveland's terms were not consecutive is irrelevant.

Because of these facts, site.net describes the current President George W. Bush as the 42nd President of the United States. Many sources refer to him as the 43rd President, even though there have only been 42 individuals that have served as President.


MrStitch

5+ Year Member



 
Msg#: 3501195 posted 10:53 pm on Nov 14, 2007 (gmt 0)

hahahha...

Thats actually a nice bit of useless knowledge. I'll have to remember that! ;)

potentialgeek

WebmasterWorld Senior Member 5+ Year Member



 
Msg#: 3501195 posted 2:53 pm on Nov 15, 2007 (gmt 0)

Bounce rate tracking gives me a bad vibe (privacy invasion). Every time I visit a site, I'm supposed to wonder if Google is checking how long I spend on it? No thanks. Spending a lot of time on a competitor's site would help it? No, but I'd still think about it, which I'd rather not do.

I've wondered lately about bounce rate and ads. If you have ads in the heat map, obviously visitors will click on them and your bounce rate goes up. Is that going to result in a hit to your SERPs?!

I hope Google is smart enough to find better ways than bounce rate to determine SERPs. Give us a flag option...

[ ] This site sucks!

lol!

p/g

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3501195 posted 3:07 pm on Nov 15, 2007 (gmt 0)

Using nofollow on internal links which in turn should stop Google from indexing all the duplicate templated content.

I have to wonder who thought of that one.

grant

10+ Year Member



 
Msg#: 3501195 posted 1:08 am on Nov 16, 2007 (gmt 0)

I was hoping to make comments earlier this week, but here they go.

I think that 'bounce rate' for Google Search is a different critter - it is 'did the searcher return from that page and click on a different result?' That's a bounce for search.

Or someone is just price comparing? I think it is very dangerous to jump to conclusions with this kind of metric.

Bad example. Adobe ranks #1 for Acrobat Reader because:...

Crap, I meant to say “ranks #1 for “click here”, which is why I said, “The most natural link is either “click here”. I can see what that might have thrown off the thread a bit.

Another flimsy example. Look again: that page is optimized for financial services.

I know. Come on, as Schleppguy pointed out, look at all the pages in the index optimized for that term. But it is the PR 10 link that enables this rank. If you don’t believe me, optimize a page for this term and see how you compare without a PR 10 link with KW in link text.

Try registering a domain for competitive keywords and building 200 links per month to that website and then get back to me on that on :P

I’ve been doing this for years. It works. When I throw a lot of money at a domain with high PR links that are topically irrelevant, the site shoots to the top ranks.

In fact, I just created a case study that shows a client their ROI on link spend against their revenue. Basically what we show is that their revenue spikes when we spike the link spend. This site is a new domain that was on page 1 for competitive terms within 90 days of site launch. (FWIW - they were spending about $3k per month on high PR text links, none were "topically relevant").

Re: the topical relevance, Schleppguy pretty much said exactly what I am saying: so are you going to argue Adobe is relevant to all of these topics? No, it's the keyword rich internal links from high PR pages.

If there are any points I am not commenting on that you would like me to, just let me know. There have been a number of comments since the OP and I am doing a quick sweep through.

Miamacs

5+ Year Member



 
Msg#: 3501195 posted 1:00 pm on Nov 16, 2007 (gmt 0)

A single site paid $3000 a month for... links...?
It took 3 months, ie. roughly $10.000 for them to reach the top?
And you come here preaching that PR is important.

Yes, PAGERANK is important, but what you paid for was TOOLBAR PAGERANK. And the fact that you can't tell topics, relevancy and trust away is the reason for the need to spend ANY money, let alone ... did I read that right? $3000 a MONTH?

Do you know why you had (have?!) to make them spend a fortune?
Because 80% of those links have NO value! They're devalued!
It's not like topical links score more points, but that irrelevant links score NONE.

And the remaining 20%, well yes, big numbers as they say.
But even if they can spend this much and their ROI is in the plus, this is money down the drain. And those links will disappear as soon as Google finds them, or they don't pay their rent. Congratulations, but this isn't SEO. This pure SEM + a little SEO knowledge from 2003.

...

And I'm having those links for free?
Don't need as many as your clients do either?
So your link-pounding tactic worked?
To put this in perspective, your SEM tactic is like a Red Army advance in WWII. While SEO is more like the SAS, CIA or a bunch of ninjas. ( little resources, don't leave a trace, no backup, act for the greater good of... *smirk* )

...

You see, here's the deal.
If you gave me that $3000, once, there wouldn't be any spikes.

Because the rankings would STICK.
Can you see the difference now between your FFA run for links and a real link 'campaign'?

Whoa. You had me for a moment there.
Yes, PR is important. Yes your tactic works.
And no, it's not the best, you can't conclude anything from it, and ... wow... $3000 a month just on links. I'm lost for words.

...

OK, enough of this, Congratulations, and I don't mean this out of sarcasm.

I'd like to have a consultation service where I can do math with numbers like this to spend on behalf of clients. And also, since your tactic - even if it really is like the Russian Army - worked. They won the war you know.

CainIV

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3501195 posted 1:03 am on Nov 17, 2007 (gmt 0)

If the 'tactic' worked then I guess it worked, and it's all about results so they say.

Still not sure I will concede or agree with your assessment that you can build 200 one way links to a new domain and expect good results.

I will however concur, as unfortunate as it is, that a handful of good paid links in the right places can make or break rankings when done right. Most webmasters probably read way too much into paid links as part of the paid link scare '07.

Even if a website is demoted for visible pagerank, that doesn't mean behind the scenes that it doesn't pass pagerank or ranking value.

jakegotmail

10+ Year Member



 
Msg#: 3501195 posted 3:47 am on Nov 17, 2007 (gmt 0)

$3000 a month just on links. I'm lost for words.

If this seems like a lot to you for links in competitive industries then you must really be oblivious.

Even if a website is demoted for visible pagerank, that doesn't mean behind the scenes that it doesn't pass pagerank or ranking value.
100% agree. You really cannot connect the two. TBPR is garbage. You can't discern that much from a PR droppage.

Miamacs

5+ Year Member



 
Msg#: 3501195 posted 1:55 pm on Nov 17, 2007 (gmt 0)

If this seems like a lot to you for links in competitive industries then you must really be oblivious.

No, it's not much on the pre-packed, off the shelf link market, I know.
But... let me be surprised just a little bit, for the OP sounded kinda SEO-test-ish ( saying 'PageRank is the only thing you'll ever need - have proof' ), while spent $3000 on monthly fees buying high PR links at random? Something to pay every month, and something that lasts only as long as it is paid for?

...

What I said was, that out of $3000, and with proper knowledge, you can do a lasting link campaign. Something that isn't a constant burden.

Something that's an investment, rather than paying back your mortgage interest... Buying links from brokers, or people who sell them openly, and/or buying links with complete disregard of active, well known, working Google filters is... a waste of resources.

Whether it'd be *targeted* bought links, money to oil the gears for a permanent 'reference / citation', pocket money to get a PR article from above, AND/OR it'd be 3 SEOs who you ( okay, *I* ) can instruct on what to do for a month, that money can be spent on something that'll last, instead of a temporary 'boost'... that's almost like using AdWords.

Yes, compared to how much my clients pay for links ( especially 'monthly rent' for which they don't ) $3000 is a lot. I know the prices, but I'd rather learn WHY those links work, and find the ever-cheaper way around. No, not spamming. Bartel, intelligence, research. A better investment.

research... for eg.: If you keep doing SEO by buying links without any other requirement than toolbar PR... and can ONLY deduct that high PR is important... well *cough* sure, but what else was measured in that 'test'? Out of bought links, with no regard for anything else... why, of course. Higher PR is better. But if you buy 10 high PR links without a check for relevancy, crawl rate, trust, titles, whatever, 6 to 8 will have little effect on your site's overall rankings. Understanding the difference will save you a lot of money so that you can spend it better. If you don't know, can't/don't want to tell the difference, YET the few working links can carry the weight, this method may still worth it. But it's by far not the most effective, while I thought SEO was *about* effectiveness. And not every business has that much to spare either.

Sorry, not sure what we're discussing anymore... that PR is all you need? No it's not.
Is it important? Sure... extremely. Leading the top 3 of all parameters used.

But...
Link quality measure for dummies(tm) = PR x relevancy x trust
If either is low ( let alone 0 ), that link is worthless.

PR is 'sugar'. Relevancy is 'vitamins'. Trust is 'no preservatives + valid expiration date'. *smirk*

grant

10+ Year Member



 
Msg#: 3501195 posted 8:21 pm on Nov 17, 2007 (gmt 0)

Yes, PAGERANK is important, but what you paid for was TOOLBAR PAGERANK.

I know that the toolbar is only an approximation. I’d like to think everyone on this thread knows that and we're beyond having this discussion.

Because 80% of those links have NO value! They're devalued!

Um, no, they worked. I don't see how you be so sure.

But even if they can spend this much and their ROI is in the plus, this is money down the drain. And those links will disappear as soon as Google finds them, or they don't pay their rent. Congratulations, but this isn't SEO. This pure SEM + a little SEO knowledge from 2003.

It is not money down the drain if you saw how much money they are making. I am talking millions of dollars every month. Spending $3k is absolutely nothing to them. It might seem a lot to you and me relative to what we keep in our checking accounts.

They could spend $10k per month on links and it’d still be a tiny, tiny expense for the return they are getting. Insofar as “not SEO” – look, I read WebmasterWorld all the time, I know what the seasoned folk here think of as best practices. With some clients, spending on links like these makes more sense than other tactics. Cain IV doesn’t believe a site can rank this quickly. Why? Because a lot of people here have never tried spending this kind of money on links. I will give you this – it is ENORMOUSLY important to choose the links wisely. You are right that there are a lot of crap links you can buy. You have to be smart about it for this approach to work.

You see, here's the deal.
If you gave me that $3000, once, there wouldn't be any spikes.

We do get permanent one way links, but they take longer. We have those campaigns going on concurrently, but the point of my OP was that these links buys work, and it is the PR of the pages, not topical relevance, that gets us there.

I can sense that many of you are thinking, “but I can do that for free”.

I report on ROI, and I always compare the cost of the SEO to PPC. Typically, I get keywords that would be $3-8 per click for about $.25 in SEO. We far exceed the goals of the campaigns, and the advantage of paid links is we can get them immediately. The amount of time it takes to get free links might be a matter of months. Considering the client makes a few million every month, how much do the free links actually cost (over time)?

Again, we have campaigns running to get the free links for the long haul, but you have to see the large economic picture here to really be in a position to make a judgment call.

And to summarize, recall my OP was about the effect of these “off-topic” links that are from high PR (let’s lose the toolbar discussion please) sites. I don't see this as black or white (paid vs. quality free), I am just saying paid links work wonderfully if you choose them well and in the context of client ROI, they can be a necessity.

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3501195 posted 8:32 pm on Nov 17, 2007 (gmt 0)

The fact that paid links do generate a healthy ROI is exactly why the paid link market is such a challenge to Google and why they kicked off their recent campaign.

It's interesting that Google has not been confident of their algorithm's ability to handle the effects that paid links bring into the search results. Their additional use of social engineering and spin around paid links has certainly added to the collection of SEO myths - but at least in this area, the final chapter of the story is not yet written.

superclown2

WebmasterWorld Senior Member 5+ Year Member Top Contributors Of The Month



 
Msg#: 3501195 posted 11:06 pm on Nov 17, 2007 (gmt 0)

Google likes to see a steady growth in links, a new site suddenly having hundreds right from the start is OK but only if the following month there are hundreds more. And then, the following month, hundreds more again. Buy 3,000 dollars worth of links this month and 3,000 dollars worth next month and you may as well put up a poster outside the Plex telling them that the links didn't come naturally. Still if there's millions a month in the pot I'll happily sell you a few thousand <G>!

annej

WebmasterWorld Senior Member annej us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3501195 posted 8:22 pm on Nov 18, 2007 (gmt 0)

We all know that Wikipedia kicks ass for nearly ever search planet on the term.
Please define their topical relevance. What is their "topic" in the eyes of Google?

I've thought about this and have done some looking at Wikipedia's structure.

Many of the links to those minor pages, including stumps, do come from related pages. For example "outdoors" points to "plants" point to "trees" and so on. Wikipedia automatically links from related pages to other related pages. Unfortunately there is no human judgement here so a stump with all kinds of criticism about the page does as well a an excellent Wikipedia entry with extensive information on the topic.

grant

10+ Year Member



 
Msg#: 3501195 posted 1:35 am on Nov 19, 2007 (gmt 0)

For example "outdoors" points to "plants" point to "trees" and so on.

You're reaching here. You found a few examples of related links, which is inevitable given their site.

For instance, look at the Wikipedia page on Al Franken.

First sentence about his life:

"Franken was born in New York City to Phoebe G. (Kunst), a homemaker and real estate agent"

"New York City" and "real estate agent" are links.

Is Al Franken "topically related" to either of those topics?

Not in the sense that you are looking for. You're looking for "outdoors", "plants", "trees". Relationships are far more complex than that, and outside of this one page on Franken, there is no "topical relationship" between Franken and real estate.

Superclown2:
you may as well put up a poster outside the Plex telling them that the links didn't come naturally

I guess you've missed the point of the thread, which is to specifically identify what makes links look "natural". Because I think for every profile you will create, there will be at least one prevalent exception. There are way too many false positives, and it is the lack of this recognition that has me believing most of the "natural" talk about links is a myth.

Also, I cited specific examples where paid links work. If you would like to present a counter point, I expect to see specific examples. Myth is fueled by hot air.

Tedster - I really like the way you contribute to threads. I can see that you resisted some of my ideas at first, but you tend to arrive at an open-minded conclusion (even if you are not 100% on board with my ideas). This approach is how one gets the most out of this board. It's good to have you here.

annej

WebmasterWorld Senior Member annej us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3501195 posted 3:50 am on Nov 19, 2007 (gmt 0)

I don't know if this maze of interlinked pages based on the fact a the two pages have a related word would be called relative links. I was thinking more of how they link pages that are loosely related. But there really is a natural relationship. Like transportation, trucks, Fords, 350 or whatever. I'm not even saying the fact they are related makes a difference. Just pointing out there are some related links that might make a difference.

But I do believe a far greater factor is just how huge the site is and how much linking power that brings. That mixed with the PR they have. But we can't be sure if there is a little extra weight on the related stuff.

FlexAjaxSEO

5+ Year Member



 
Msg#: 3501195 posted 4:29 am on Nov 19, 2007 (gmt 0)

I have been on this site for a good 5 years, changing my name every 5 months to avoid having too much data in one place for google to track my paranoid comments. Google watches this forum closely and SEO types should beware, seriously.

CainIV

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3501195 posted 4:46 am on Nov 19, 2007 (gmt 0)

Google watches this forum closely and SEO types should beware, seriously.

Possibly, but how would that change the fact that Google apparently has no [current] way to accurately calculate how natural a link is is not.

For instance, look at the Wikipedia page on Al Franken.
First sentence about his life:
"Franken was born in New York City to Phoebe G. (Kunst), a homemaker and real estate agent"
"New York City" and "real estate agent" are links.
Is Al Franken "topically related" to either of those topics?

Again I think using the example of Wikipedia makes no sense here as did the Adobe site, because Wikipedia is an authority and the rules are different for these types of sites. You already alluded to this in your original post by essentially defining that Google treats authority websites differently than most regular websites:

But one could argue these are from “authoritative” sites and won't raise a flag. Ok, think of all the esoteric niches that have the equivalent of a major announcement happen.

Nonetheless, I would concur with much of what you have said - currently, Google does not seem to have a grasp on what is or is not natural, nor does Google have any way to seemingly place correct credit where it is due in terms of links.

Miamacs

5+ Year Member



 
Msg#: 3501195 posted 1:54 pm on Nov 19, 2007 (gmt 0)

Grant are you entirely sure you know what topical relevancy / themes are? I mean from the standpoint of Google...

Okay...

The Theme(s) of a website isn't (aren't) limited. They're not exclusive.
They're inclusive.

In short, consider them licences to add a new page on the site...
With a relevant, competitive, popular anchor text...
...and not get filtered out or sent back to the -950 fun land.

The licences are granted based on the incoming links' anchor text, their source, and not based on the current on-page content!

...

To those who don't get it:

Wikipedia has thematic relevancy to nearly everything.
Not just because its massive content on everything, and not only because their internal navigation ( which is world class ), these alone wouldn't be enough. However, its incoming links support all these themes... which relevancy then gets carried around by the (proper) navigation, to the (proper) content... thus each page is soaking in "relevancy" to many, many *other* things. They have millions of inbounds supporting their wide range of topics, sometimes down to competitive phrases.

Same goes to news sites.
Universities.
Review sites.
Shopping sites.
eBay, Amazon, and... Adobe.

There are a *lot* of ( high PR ) sites you'll never see appear on the SERPs unless you look for them by their maiden name. Even though they may as well have some pages dedicated to stuff you are looking for. Do you know why?

Those listed above have the same amount of incoming links, but with a wide variety of anchor text, to a wide variety of subpages, supporting many if not all of their themes, even if they're so numerous you've never crossed their ways before. A site with 100.000.000 links saying 'click here' *only* would NOT rank for Financial Services. Yes, a site can have multiple themes if it can support these, or rather, if it can get support from its inbound links' anchor text. And if it can match this up with a navigation that builds a proper relevancy-pyramid, no matter where the links point to within the site, the entire domain will benefit from it. The 'licence' to post related material, use competitive wording isn't revoked when the content disappears. It goes into slumber mode, making the page relevant but not appear on the SERPs unless it really has the stuff on there.

The sites you bought links from may not *look* topically relevant to you ( I wish I could see that case study, with the effects broken down link by link ), and yet very well be carrying latent relevancy to many other things than which were apparently not on-page at the time. On the other hand, the NYT wouldn't have had to publish a Cruise related article to be relevant to a 'Sea Cruise' link. ( are they still selling links? *grin* ) But that doesn't mean relevancy wasn't a factor. It means the requirements have been met at a much higher level already. Looking at the link profile of the one providing the link is routine.

Do you remember when Google told everyone it had fixed the Google bomb exploit? Yeah, they fixed it allright, a site linked to with a lot of stuff that isn't even on their page, won't show up in the SERPs for that phrase. Then the White House website accidentally published both words ( 'miserable' and 'failure' ) on its homepage in a news conference report. And suddenly, they were BACK at #1 for...

Meaning there's an awful lot to a site's (latent) relevancy that the naked eye won't see, unless the SEO digs deeper, and looks at a couple of dozen incoming links to the site. Or a full report, neatly sorted.

But when you deal with news or otherwise generic-themed sites for example, that's not even necessary.

...

And before you answer, please consider that I'm bringing up the technical details as to *whyyyyy* your pushy link campaign might have worked... because of being so full of *envy*! Or if you had realized already, go easy on me pleeze. *groan*

...

annej

WebmasterWorld Senior Member annej us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3501195 posted 12:35 am on Nov 20, 2007 (gmt 0)

Wikipedia has thematic relevancy to nearly everything.
Not just because its massive content on everything, and not only because their internal navigation ( which is world class ), these alone wouldn't be enough. However, its incoming links support all these themes... which relevancy then gets carried around by the (proper) navigation, to the (proper) content... thus each page is soaking in "relevancy" to many, many *other* things. They have millions of inbounds supporting their wide range of topics, sometimes down to competitive phrases.

Interesting. But some specific themes in Wikipedia have no inbound links. Am I right in thinking that may not matter because there are enough inbound links to related pages?

This 56 message thread spans 2 pages: < < 56 ( 1 [2]
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google SEO News and Discussion
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved