| 10:28 pm on Apr 29, 2005 (gmt 0)|
what do you mean with random?
| 10:47 pm on Apr 29, 2005 (gmt 0)|
Before the recent Google update algo, the link anchor had a great importance in SERPs.
Now that G is penalizing excessive repeated link anchor text (they even dropped paypal and was fixed with manual submission!), adding an "almost" random factor to the link text sounds like a good idea.
| 6:23 am on May 2, 2005 (gmt 0)|
Random anchor text is definitely the way to go. I recently was working on a site that Google didn't seem to like too much. After some further checking, it was found that too many links were obtained with the same anchor text in a short period of time. It seems that Google viewed it as an unnatural linking pattern and penalized it. Once the anchor text was varied, the old links seasoned a bit and new links were obatined in a more natural fashion - things started to improve.
| 1:50 pm on May 2, 2005 (gmt 0)|
I've been randomizing certain anchor text for a long time now in anticipation of this penalty, and it has worked to good effect. Next up for Google is to penalize anchor texts that are *too* random, or to use the current buzzword, *unnaturally* random.
I'm sticking with a few stock variations on anchor that I will rotate through.
| 2:05 pm on May 2, 2005 (gmt 0)|
<I'm sticking with a few stock variations on anchor that I will rotate through.>
Are you "gaming" Google ;-)
| 2:32 pm on May 2, 2005 (gmt 0)|
This thread approves one more time that google’s page rank according to IBLs is bad idea.
Artificially inflated number of IBLs mistakes Gbot and reduces the quality of G search results.
There are enough “clever” and “smart” boys that use this G’s drawback.
So, old “classic” information retrieval approach used by Yahoo looks to be better.
| 2:46 pm on May 2, 2005 (gmt 0)|
G penalizes the entire site right, not just that keyword search?
what does one do about rss feeds if sites run them sitewide?
| 4:59 am on May 4, 2005 (gmt 0)|
what is the verdict on using random anchor text for site's own internal links?
| 5:28 am on May 4, 2005 (gmt 0)|
I was wondering the same.
It makes perfect sense to have the same anchor text throughout a site on nav bars for consistency, but for variety I could see having alternate text scattered throughout the site, in places other than nav bars.
But this is becoming bizzaro world, where everything I once knew is often the opposite now.
| 6:18 am on May 4, 2005 (gmt 0)|
This may be a naive question, but why not simply do what's natural instead of trying to make things look natural?
| 7:43 am on May 4, 2005 (gmt 0)|
Google's patent talks about this randomizing nonsense. Doing anti-natural things trying to pretend to be natural is too mind-boggling to even think about, so it would make sense to be one of the blackest marks you can attract against your sites.
| 8:04 am on May 4, 2005 (gmt 0)|
Even from a white hat perspective it's frustrating to second guess what Google considers legitimate and what they don't like to see in terms of inbound links and the anchor text associated with those links.
For example if a travel site runs a "keyword hotels" link as a sponsor on home page nobody would say it's a spammy link, but how about on every page of a large site? Half the pages? Drug links? I think the Google Guidelines are far too vague with respect to legitimate linking practices.
For a complex site many anchor text terms are legitimate, but anchor text abuse is a common spam technique. (The guys at SEOinc would probably like clarification of this as well!)
It seems to me randomizing relevant anchor text terms is a good example of something that might be considered out of the spirit of the Google Guidelines but probably works well since it would appear as relevant inbound links from many pages.
| 8:13 am on May 4, 2005 (gmt 0)|
Inbound links are a nightmare of information retrieval theory.
It should be eliminated as a bad example of relevance ranking.
Link farms, reciprocal links, IBLs artificially inflated by webmasters, aren’t it a result of G’s page rank approach?
| 4:41 pm on May 4, 2005 (gmt 0)|
When did Google drop PayPal? Was this documented anywhere?
| 5:05 pm on May 4, 2005 (gmt 0)|
paypal's story is sorta funny, for some quite relevant KWs (perhaps many) they dont have the top G serp.
| 5:25 pm on May 4, 2005 (gmt 0)|
the fact is most natural links are the url posted whole and no anchor text at all. Sites wihth no inbound links as pure urls are easy to spot as non-organic.
| 5:38 pm on May 4, 2005 (gmt 0)|
|the fact is most natural links are the url posted whole and no anchor text at all. |
That might have been true years ago, but I don't think it's true today. And it goes against one of the fundamental principles of the Web: namely, hypertext linking. I can't imagine Google thinking there's anything "unnatural" about a link that says "National Widget Corporation" and takes the user to nationalwidget.com. OTOH, Google just might think there's something unnatural about 50 or 500 or 5,000 links to an e-commerce or affiliate page with artificially randomized anchor text like "free widgets," "cheap widgets," and "discount widgets." :-)
| 5:48 pm on May 4, 2005 (gmt 0)|
The best reason to vary anchor text is that different text could attract different clicks.
With internal links using varied anchor text can increase the number of visitors that click through to a given page. That works well when you have more than one link to a page on any other given page. For instance a link in a standard site navigation block and a link in the text on the page.
With external links there's probaly nothing wrong, and quite likely some benefit, with making the anchor text appropriate to the page or site the link is placed on.
| 5:52 pm on May 4, 2005 (gmt 0)|
I find it very disheartening that it's an issue to worry about.
My site has a unique name. I haven't actively solicited many third party links and certainly no links from irrelevant sites or run of site stuff etc. The site gets the vast majority of incoming links naturally because it's the top resource in its field. Just about everyone links to it using the title as the link text.
Now I find myself on page 3 of the Google SERPS if you do a search on the exact site title. Ranking above it are a bunch of pages that link to my site using its title. Some of them are scraper directories.
Hard to interpret that as a positive for the searcher.
| 5:58 pm on May 4, 2005 (gmt 0)|
|For instance a link in a standard site navigation block and a link in the text on the page. |
hhm thats interesting. i think Yahoo used to specially favour that ...
| 6:01 pm on May 4, 2005 (gmt 0)|
|That might have been true years ago, but I don't think it's true today |
its certainly true for all the sites i check. Im not saying no links have anchor im saying a standard percentage are naked urls and if a site has 100's or 1000's of inbounds and the percentage of naked urls is not inline you can guarantee they are non-organic links.
| 6:12 pm on May 4, 2005 (gmt 0)|
not naked url, but domain.com or domain, like Verizon or Google. linked to your URL. Google apparently doesn't like that either for smaller sites.
| 8:43 pm on May 4, 2005 (gmt 0)|
by naked i mean