Backlink Analysis Strategy

[achronister]

Hi all. Right now I am working on doing a full analysis of my backlinks and I'm curious what everyone thinks about this scenario.

I can get the following data: linking url, linking domain, linked to url, and anchor text.

What I'm wondering is if you would base your anchor text percentages on linking domain or linking url?

(example: I have a few sites with sitewide links. Would you count the entire domain as one, or would you assign weight to each individual page?)

Aaron

[brian.mark]

I'd only count one per domain, unless you're looking at this from an MSN perspective. They are the only ones that seem to give you any form of link credit for every link on a site (such as forum signatures.)

Brian.

[digitallabz]

I'd only count one per domain, unless you're looking at this from an MSN perspective. They are the only ones that seem to give you any form of link credit for every link on a site (such as forum signatures.)


I agree with Brian.

Article Submission (http://www.articlenetworks.com)
Powerpoint Templates (http://www.digitallabz.com/powerpoint-background.html)
Submit Articles (http://www.articlenetworks.com)

[Chris]

good to see you aaron. missed you in chicago.

i agree w/counting the one domain as well.

[achronister]

I'll go with the one per domain and see what I come up with. It should be easier on the crawler I wrote to parse the links anyway. Thanks for the feedback.

Chris - I'll be in NY in Feb for that one, we'll grab some beers this time.

[ADAM Web Design]

I'd only count one per domain, unless you're looking at this from an MSN perspective. They are the only ones that seem to give you any form of link credit for every link on a site (such as forum signatures.)

Brian.

http://search.sympatico.msn.ca/results.aspx?FORM=TOOLBR&q=Toronto+web+design

Keep it down, dude. That's working for some of us. ;)

Seriously, I too am finding 1 per domain seems to be about the limit for big G.

[sandis.viksna]

I disagry a bit,

First of all, it depends does the links goes to your home page, or your internal pages as well; if all links to your domain, then it is not too good since of dupe anchor text, but if your website gots bunch of different links with different texts to different parts of your site and is within the scope (subject), then this makes lots of sense and relevance!

Second thing, PR... is on-page factor that is being splited between links that are on the exact page so it should five you some bits of this too, especially if linked to different parts of site;

[sem-seo-pro]

Hi

I would have to say studying backlinks is a waste of time. Ssince nobody has a clue as to which links Google is going to give which value, what exactly are you studying?

Next the links were built in the past...even if it was an hour ago it's still in the past.. and as we all know you cannot go back and replicate the past.

Next while you are studying the past in a way... the present is reshaping... and changing the future.. tomorrow is the today you dreamt about yesterday..news happens, pages fall offline, servers crash, new sites are added, new pages are added, new links are built, others fall offline and all of which then changes the mathematics applied at the time to calculate the score, and since all of these factors are instant at the time of occurance, the scoring is impossible to determine exactly, and thereby leaves us to best guestimates....

To the extent that anyone could use the data to move pages to the front of Googles listings, by the time all the factors were considered, things would have changed again.

Dog chasing tail...??

There was a time the Internet could be best described as an animate static item that could be replicated for results, however those days are gone.. the Internet of today is alive as much as any of us are....what is there now will change soon.. it is invetable.. and all the supposition in the world won't change that.

Reciprocal links are dead.. would you agree?

So the only real valued link is a one way link??

Given Google has told us links age and lose value, then how do you determine the depreciation factor??

(Sounds IRS like.. ugh :-| .. )

Seems the best thing to do is produce quality content that sites will want to link to by themselves, as this is how the web was built, and how Google seems to be viewing things based on Historical Data Retrieval.

Happy New Year!

Clint

[achronister]

I would have to say studying backlinks is a waste of time. Ssince nobody has a clue as to which links Google is going to give which value, what exactly are you studying?

I'm using the Yahoo Site explorer API


Next the links were built in the past...even if it was an hour ago it's still in the past.. and as we all know you cannot go back and replicate the past.

I can use this data going forward



Reciprocal links are dead.. would you agree?

Agreed, but irrelevant. I didn't say I was looking at reciprocal linking partners, I am looking everything.


Given Google has told us links age and lose value, then how do you determine the depreciation factor??

Since when?? That's a whole different argument anyway, and I certainly don't recall Google telling anyone that old links count for less.

The goal of this exercise is to look at the total inbound links for a certain page on my site, and find out what anchor text is being used for each link. Say for instance I have a page with 600 inbound links, and 12% of them are using the same anchor text. I don't need to worry about this one as much as a second page with 75% identical anchor text. The point is when I go out to secure new links, I will know when and how to alternate my anchor text. If you don't think you can use data like that, I'd have to disagree.


Seems the best thing to do is produce quality content that sites will want to link to by themselves

Sure, that is always a goal. In markets like mine, and yours presumably, this strategy alone will not work to achieve the desired results. Linking currently is, and I believe will be very important in 2006. Linking strategies will have to become all the more creative.

Aaron

[amabaie]

Say for instance I have a page with 600 inbound links, and 12% of them are using the same anchor text. I don't need to worry about this one as much as a second page with 75% identical anchor text.

In either situation, you have obviously not been esuring enough variety. I would spend less time analyzing, and just keep doing what makes sense to do, regardless of the history.

Yes, sem-seo-pro makes some pretty outrageous claims, but the core message is sound. My personal opinion is to forget analysis and do what makes the most sense on its own.

[sem-seo-pro]

Hi Aaron

You can google search 'google patent march 2005' to get the full content but I will post here the part that is important and directly related to your post.


[0066] Link-Based Criteria

[0067] According to an implementation consistent with the principles of the invention, one or more link-based factors may be used to generate (or alter) a score associated with a document. In one implementation, the link-based factors may relate to the dates that new links appear to a document and that existing links disappear. The appearance date of a link may be the first date that search engine 125 finds the link or the date of the document that contains the link (e.g., the date that the document was found with the link or the date that it was last updated). The disappearance date of a link may be the first date that the document containing the link either dropped the link or disappeared itself.

[0068] These dates may be determined by search engine 125 during a crawl or index update operation. Using this date as a reference, search engine 125 may then monitor the time-varying behavior of links to the document, such as when links appear or disappear, the rate at which links appear or disappear over time, how many links appear or disappear during a given time period, whether there is trend toward appearance of new links versus disappearance of existing links to the document, etc.

[0069] Using the time-varying behavior of links to (and/or from) a document, search engine 125 may score the document accordingly. For example, a downward trend in the number or rate of new links (e.g., based on a comparison of the number or rate of new links in a recent time period versus an older time period) over time could signal to search engine 125 that a document is stale, in which case search engine 125 may decrease the document's score. Conversely, an upward trend may signal a "fresh" document (e.g., a document whose content is fresh--recently created or updated) that might be considered more relevant, depending on the particular situation and implementation.

[0070] By analyzing the change in the number or rate of increase/decrease of back links to a document (or page) over time, search engine 125 may derive a valuable signal of how fresh the document is. For example, if such analysis is reflected by a curve that is dropping off, this may signal that the document may be stale (e.g., no longer updated, diminished in importance, superceded by another document, etc.).

[0071] According to one implementation, the analysis may depend on the number of new links to a document. For example, search engine 125 may monitor the number of new links to a document in the last n days compared to the number of new links since the document was first found. Alternatively, search engine 125 may determine the oldest age of the most recent y % of links compared to the age of the first link found.


Aaron I read your replies and found them very well presented, and hope the informaton and thoughts I have given you, will help as well.

Clint

[achronister]

In either situation, you have obviously not been esuring enough variety. I would spend less time analyzing, and just keep doing what makes sense to do, regardless of the history.

Those were theoretical numbers. I have no idea what they really are; I was just explaining what I was doing.


forget analysis and do what makes the most sense on its own

Just wing it in other words? With over 80,000 inbound links from external domains I think I will see where I stand, identify possible problem areas and continue from there.

[sem-seo-pro]

amabaie said

Yes, sem-seo-pro makes some pretty outrageous claims, but the core message is sound. My personal opinion is to forget analysis and do what makes the most sense on its own.


Thank you!!! :)

Happy New Year!!

Clint

[sem-seo-pro]

achronister said

With over 80,000 inbound links from external domains I think I will see where I stand, identify possible problem areas and continue from there.

Hmmm tell me please, in my eyes, with 80,000 inbound links the % of traffic from such,.. should generate close to 85 to 90% of your traffic, if this is true as is often the case...then why worry of the 10% or so traffic from SERPs?, especially when its obviuos people like your site for other reasons (good content?) than a pretty search engine listing.


Clint

[achronister]

sem-seo-pro

You make a solid point on the date relevance, however I still think there is a lot of assumption going on. There are a lot of 'may, might, could, in one implementation, etc' going on in that patent paper.

Back to my main topic...I think it is better to know where I am with regards to the anchor text of my inbound links. This way I can avoid making any mistakes by overuse of certain phrases, etc. I could randomly guess and just wing it, but that doesn't make much sense to me.


with 80,000 inbound links the % of traffic from such,.. should generate close to 85 to 90% of your traffic, if this is true then why worry of the 10% or so traffic from SERPs

Where are you coming up with those numbers?? Is that a guess or based on some knowledge of my traffic numbers and/or traffic numbers in my niche? Is there some correlation between number of links and percentage of traffic from said links?

Why worry about the SERPs? If everyone were content with the status quo, there wouldn't be very many successful organizations.

[sandis.viksna]

in my industry, there isn't any free traffic links, just these one in deep corners of websites, and traffif definatelly doesn't comes from these farest corners.

I agree, the SE traffic is the one we need to work worfard;

/sorry for posting this, but i agree with this guy who wants analize not just wing/

[sem-seo-pro]

achronister said

Where are you coming up with those numbers?? Is that a guess or based on some knowledge of my traffic numbers and/or traffic numbers in my niche? Is there some correlation between number of links and percentage of traffic from said links?

No I do not know your site or it's traffic numbers. This is merely from what I have seen with other sites with such high external inbound links, and there are a few other posts here on WPW with the same basic data and thoughts.

achronister said


Why worry about the SERPs? If everyone were content with the status quo, there wouldn't be very many successful organizations.

Very true, well said.

Peace

[sem-seo-pro]

in my industry, there isn't any free traffic links, just these one in deep corners of websites, and traffif definatelly doesn't comes from these farest corners.

I agree, the SE traffic is the one we need to work worfard;

/sorry for posting this, but i agree with this guy who wants analize not just wing/


No need to be sorry.

I think you bring a valid point to the post that there are always exceptions and this is what we need to learn.

Just as a realtors website would have a lower bookmark percentage than say a daily joke website.

The thing with the Internet is it is so new and so diverse, so much so that there is very little that is similar to compare site to site side by side.

Happy New Year to you

[dburdon]

I would have though that for any link to have real value it should generate traffic as well as PR.

I probably generate 4 times as much traffic from links as from the SERPs. Am I the exception?

[achronister]

I would have though that for any link to have real value it should generate traffic as well as PR.

That is the goal, and what to look for when digging for links. As for the traffic numbers, not really sure if you're the exception or not. We pull in more from search than link referrals.

[sem-seo-pro]

I probably generate 4 times as much traffic from links as from the SERPs. Am I the exception?

Not from what I have seen, but again, each site at times is an oasis unto itself.

Clint

[KennyBP]

OK, given that all except MSN count 1 link per domain, which would you give more weight to:

- PR and being higher up a heirrarchy (generally top level pages have better PR from my experience) OR
- Page relevancy and being on topic.

Or would a mixture of the two be best?

[ADAM Web Design]

Basically, my attitude on IBLs is one rule:

Get as many free one-way links as you can. If a link isn't paid for, there's no way that I can be faulted for it, since anyone can generate a free link from anywhere.

Even if the link doesn't have any SEO impact, if it's placed in a decent location it should have a direct traffic impact.

How does the saying go? "I don't care if it's good publicity or bad, just spell my name right."

[sem-seo-pro]

Adam

I think it was "bad publicity is still good publicity"

How else would you explain Martha Stewarts jail term success?

Happy New Year

[Mel]

The assumption that any of the points in the Google patent mentioned above are actually being implemented has yet to be demonstrated in any way, and until such time as it has been, I would caution that such assumptions are just assumptions with not much to bqack them up, and certainly not proof.

If you read that patent carefully you will find sections that state that traffic affects rankings etc etc.

I believe that patent was applied for simply to put a hold on any areas which Google might like to research in the future, not as an indication that it is something which has been actually put into practice.

I have also seen some suggestions that this patent may have been a red herring to lure naive webmasters into thinking that this was the Rosetta stone to the Google algo and do not totally discount that view.

[sem-seo-pro]

Mel said
I would caution that such assumptions are just assumptions with not much to bqack them up, and certainly not proof.


I would caution, that assuming there were assumptions made, when there were none, this may lead others to look at the poster for what his true intentions are.

And as far as the patent goes there is a lot of proof that by my following it I find sucess.

What you do Mel,, is your style.. Yours to keep.. but certainly not an indicator that your methods are right or wrong.

And as far as traffic to a website...it is indeed a criteria needed, in order to move up in the SERPs, a site with no traffic to Google, is a site with no traffic. Not very popular and not one I would recommend friends (users) to.

Ask yourself this question ....if there are 100 websites in any category, then why would Google send traffic to a brand new website #101??

Lastly this statement is hard to believe you even posted.


that this patent may have been a red herring to lure naive webmasters into thinking that this was the Rosetta stone to the Google algo

Where did you make that up from lol???

Yes lets waste millions of dollars on Patent Research to trick seos and webmaster... lol...lol

Best Laugh of 2006 and its only just started.

lol

[williamc]

that this patent may have been a red herring to lure naive webmasters into thinking that this was the Rosetta stone to the Google algo

Where did you make that up from lol???

Yes lets waste millions of dollars on Patent Research to trick seos and webmaster... lol...lol

Best Laugh of 2006 and its only just started.


Google has hundreds if not thousands of patents. They are simply making sure that ideas they may think will be useful down the road are covered and owned by them intellectually. Many companies patent ideas for this very reason. Many companies, including google have been said many times to purchase patents because they realize that some webmasters who just do not know better will think they are the rosetta stone as Mel put it and blindly go around telling people that recipricol links no longer count, or that google is using clicks on listings to rank websites.

This is nothing new and is just good business practice for google to do, as it makes people look to other areas than they really should be looking. This makes their algo harder to crack and get top rankings for.

This is nothing new and Mel is hardly the first to think it is possible.

[brian.mark]

Actually, Dr. E. Garcia (http://www.jupiterevents.com/sew/summer05/garcia.html) (if you don't know who he is, you're not into search engines as much as you think you are) had a nice presentation at SES in San Jose (http://www.jupiterevents.com/sew/summer05/agenda.html#900-2) about those patents. He basically said that while nice on the surface, there is no real way to implement those technologies at the moment. They may be factored in at some point in the future, but for now they have the patents as a sort of "Wish list" for future algo's.

That said, he thought a few of the concepts were being worked on but not yet in practice. Granted, this was August, but many of those concepts aren't a simple "add it into the mix right now" type of an item. They'd take a lot of development an implementation.

For some of it, Google Analytics may be starting to set the foundation. If they find sites where people engage, those sites must have value. It could make spamming a thing of the past, but they're not to that point yet. They'll need much more participation to see any real value from that data, and their servers can't handle it yet.

It was also pointed out several times at that session that these are papers made public and just say the types of technologies that they'd like to use, not how they're going to be used. They'll never make the algo a matter of public record.

Publicly published items won't ever disclose any real information, but just a generalized overview of things they might take into account someday or that have already been in practice for a while (patents take some time to get through the process.) That's the main reason you can't read about a patent and then expect it to be in use the next day / week / month like some webmasters expected when the patents came out. They were written some time back, so some may be in use, some may be on the scope, and some parts of them may be a smokescreen. Usually all three are included in Google's patents any more.

Brian.

[williamc]

Nice post Brian. Orion is generally spot on with my own lines of thinking except in a few minor cases, and he definately knows his stuff.