Lies…
Notice google.com backlink ratio between Google and Yahoo – Yahoo figure is nearly 163 times larger than Google’s own figure! And interestingly www.amazon.com is returning " Your search - link:www.amazon.com - did not match any documents.".
Damned lies….
Notice amazon.de – this time the backlink ratio between Google and Yahoo figures is amazing 1:1830. It is a well known fact that Google shows only a sampling, but selecting only 1 out of each approx. 2000 pages MUST BE fine art. I wonder what kind of algo Google uses to filter visible backlinks?
…and nothing but lies
Notice amazon.fr – Google says no backlinks, Yahoo finds only 719, AllTheWeb 485 backlinks - MSN whooping 103k. Interestingly the w/w subdomain count for MSN is exactly the same, which might indicate how MSN counts backlinks figures.
And the truth? IMHO the really trustworthy backlink information can be analyzed only from log/referer data. Everything else is more or less a numbers game where the truth is in the eye of the beholder.
As always, feel free to share your thoughts and comments.
I consider backlinks as another indicator - not as a definitive measure of success or even SEO worthiness. I pay attention to backlinks but don't freak out if there is some disparity.
With Amazon.com, who has a huge affiliate network, it's not surprising to me to see such a huge flucuation. Additionally, I think it would make sense for the search engines to test on their different regional search engines which algorithm of backlinking would work best.
As for whether it's trustworthy, I think we just need to play with the cards we're dealt.
In my opinion ... Yahoo is the one to use ... just see the first 20 links. Other 100 pages are either not as important or simply garbage.
Now ... this is based on our specific industry. In other industries it can be the first 10 or first 40.
I personally feel bad for SEOs who try to get links from all 5,000 places their competitors got links. Poor bastards :-)
Im going to go out on a limb here, but I disagree about several points...
First as a corrective measure, link:www.amazon.com in Yahoo now shows 20,583,117 results. So there must have been some issue that has since been corrected.
Secondly, log referrer data is not more valuable because it will (1) show referrers from non spiderable links (drop down menus, javascript links, flash links etc.) and (2) not show referrers from valuable links that are not easily accessible
Third, the most important part that the link counts are internally valid. If your site has 1000 backlinks and yahoo shows 800, while your competitors site has 10000 links and yahoo shows 8000, you have lost nothing in your comparative analysis methods.
I believe the real problem is not that individuals use link counts in Yahoo for SEO reporting, but simply that they use it the wrong way. These values are relative and should only be considered in that context.
I agree with link counts being mainly usefull in relative context, but are the relative values any more trustworthy than looking at real figures if they contain false/missing information as their source?
For a reason.... With small/medium sites exclusion of even one site from index (and other situations/errors) can make huge percentual differences. With very large datasets, the effect errors etc. issues have is much much lower (a common statistical behaviour).
A classic mood swing the way G can do it;)
The question is - is this kind of data trustworthy or even usefull for SEO?
IMO It really depends what kind of backlinks site has. If links have been build carefully (i.e. there are no "low value/traffic links") or they are mainly organic, then this method is very accurate.
I don't think so. But, I still look at my backlink count on Yahoo! Site Explorer. :P
I totally agree on that.
It's the link pattern that matters here. If you build a few natural(-looking) links, then it is far better than building many in a un-natural way.
In other words, it's *link pattern* of your site, that can either help or hurt it, to a great extent. IMO
I'm currently doing a week by week report on the stats of my blog. I'll give a summary below...
Google PR: 0 - 0 - 0 Indexed Pages (Google): 2 - 1 - 4 Incoming Links (Google): 0 - 0 - 0 Indexed Pages (Yahoo!): 1 - 7 - 17 Incoming Links (Yahoo!): 35 - 74 - 157 Indexed Pages (MSN): 0 - 0 - 0 Incoming Links (MSN): 0 - 0 - 1 Indexed Pages (Technorati): 4 - 5 - 5
Stats are for the last three weeks. Yahoo! loves me, everyone else thinks I suck.
Nice post as always, 2k. Well, I have a few things that I would like to share here.
Firstly
Your stats are based on extremely large websites only. So they are biased towards larger websites and cannot give accurate information for the small and mid sized ones.
Secondly and sadly,
https://www.google.com/search?q=link%3awww%2ea...is returning 2,680,000 pages linking to www.amazon.com
And…
https://www.google.com/search?q=link%3Aamazon....is returning 200 pages linking to amazon.fr
If I am not mistaken,
Search engines use estimation and probability logic when dealing with extremely large websites data like amazon, ebay and so on.
So I’d say,
If we include small and mid sized websites as well and use averages, then we might get a clearer picture of who is lying (or making an incorrect estimate) about backlink count and to what extent.
Just my 2 cents!
P.S. Analyzing our backlinks ONLY from our log/referrers data will not be any different than the above. Or will it be? Because, in this day and age, links are obtained mostly for PR boost and ranking achievement not for traffic generation. IMHO. :)