link anchor text

random yes, but does that filter it?

[soapystar]

A lot of people support the idea that its good to keep your anchor text a little bit random, or at least mix them up a bit, but i wonder if this also has a filtering effect? In other words as the number of different words used in different inbound links increases, perhaps the reduction in keyword density lowers the value of the anchor text?

Any thoughts?

[dirkz]

Do you mean that 50 anchor texts "keyword1" and 50 anchor texts "keyword2" are not the same as 100 anchors "keyword1" (PR of the linking pages all the same) in terms for ranking high for "keyword1"?

[soapystar]

i dont quite get what you just said.

i think your just talking about pr. I so then yes the pr is exactly the same. But its not just about pr, its about the anchor text and how its weighted. So i'm suggesting that 50 links saying keyword randomword1 and 50 links saying keyword randomword2 will rank lower than 100 links saying keyword randomword1 because the second random word has filtered the total count. Im not even saying this happens, im posing the question.

[dirkz]

No, I wasn't talking about pure PR, though I think it is brought in as a weight ("keyword" from a PR7 or probably much more worth than from a PR3). For this discussion it helps to assume the linking pages are at even PRs.

Now I got what you meant. I can't imagine this to happen, from an algorithmic point of view, but have no proof. Why should Google take words into account that are near the keyword in the anchor text?

[soapystar]

im talking about the actual link itself. Poeple suggest making sure your links dont look perfect and so mix up the anchors a bit. It would also make sense that google whose main core is the ranking of inbound links would apply heavy algos to links text just as much as it would to a real page, though obviuosly the calculation would be different. If i was making an algo it would certainly have to take into account the complete spread of words across all links.

[dirkz]

On the page itself there is word density for example. You think that possibly this is used with anchor text too? Funny thought, but could be :)

[soapystar]

well i would assume so, nearby text is certainly taken into account, the discussion normally is how many words.

[cabbie]

I think yes it can have a watering down or "filtering" effect.It could possibly be dependent on how high your pr is to how many different anchor text keywords you can target with good effect.

[dpplgngr]

the same link text across hundreds of links seems to work quite well for the competiton.

[soapystar]

"the same link text across hundreds of links seems to work quite well for the competiton. "

this is my point. So its not a clear cut case of "dont have perfect links, mix them up a bit", as some people suggest.

[claus]

At the moment even traditional paid link advertising (sponsorships) seems to have effect. That's one thing you definitely don't need an SEO to do for you, all it takes is an advertising budget.

/claus

[glengara]

I think the rationale was to try to make links look more "organic" and to camouflage reciprocal links in case they're next for the chop.
If all links use the same optimised link text, it does suggest an organised reciprocal campaign.