Google PageRank and Alexa traffic Info

Will Alexa data be used in the PR algo?

[ronin]

Brett recently indicated that there is increased co-operation between Google and Amazon [webmasterworld.com].

The aim of Google is to find the sites that are most popular among the web community. Backlinks indicate a popular site because other webmasters liked the site so much they went to the trouble of linking to it.

But traffic stats indicate a popular site because lots of people are visiting it - ie. high traffic is not just an indication of a popular site, it's proof the site is popular.

A site that used to be popular but is no longer will still have lots of backlinks... it will still show up as popular in any algo which depends on backlinks (and consequentially will retain residual popularity because of continuing high rankings in Google)... but if the overall traffic has dropped away regardless of numbers of backlinks, surely the site isn't so popular anymore and shouldn't necessarily keep appearing so high up in the SERPS?

Combined use of traffic stats with backlinks will help to stem the continual rise of 'link inflation'.

So... is it possible that Google might be planning to incorporate data from Alexa into its algo? What does everyone think?

[QNetwork]

I'd doubt it because Alexa's captures data only from people who uses Alexa toolbar. Alexa's results are very much skewed because not too many people uses them. Some folks also catagorize Alexa toolbar in the same group as adwares/spywares.

[Shoestring]

I for one certainly hope not, not Alexa. I don't trust their stats as being an accurate guage overall. Nope, not at all. Granted, Brett says they are fairly spot on for extremely high traffic sites, but he also notes that in years past when the traffic was less here that they were clearly off the mark. I agree 100%. Our stats with them (as well as the competitors I check) are all over the place month to month while I see no such dramatic flux in reality with our site. The old direct hit seemed to peg us about right, however. An in-house, accurate traffic rank of some sort being factored in would be welcome, but please not Alexa.

[ronin]

Okay, I didn't expect such venom against Alexa... but whether the trafic stats come from Alexa or not, that's not really my question.

Is 'an in-house, accurate traffic rank' likely to be factored in to the algo? Because my contention is without such a measure, link inflation will never end and in ten years time, if the web as we know it still exists, new sites with small budgets won't stand a chance.

[Chris_R]

Google has its own toolbar.

They would use this instead.

I think they will at some point - if they don't already.

[chiyo]

Chris_R, as usual in his enormously useful posts, is right. Google have far more data, but its not as public.

Alexa's data is far less representative, and though the consensus is that it may provide reasonable ratings for the top say 1,000 sites if you downgrade the Korean sites which have a demographic skew, it gets very flakey from there.

Google's key utility is to find content that you may not usually find - far more than the first 1,000 sites. Using Alexa data would kill it. Their toolbar provides them far superior intelligence, as far as toolbars ever can. I presume that toolbar usage has a strong skew towards webmasters, technical people who spend a fair bit of time on the web and like widgets, and people with powerful enough workstations and connection speeds to be able to use toolbars without noticeable degradation to performance.

I would not want to see Google rankings to be influenced by the browsing habits of such a group. I would prefer to see them influenced by the browsing behaviour of casual, "everyman" type users.