Referrer tracking and links

Will referrer tracking tokens affect Google link assessment

[simonv]

We want to use referrer tokens on the end of URLs to track visitors.
We have devised a method to make sure we do not offer tokens to search engine spiders.
But if we get cross-links from other NON-search engine sites, most will simply use the URL INCLUDING their unique referrer token.
My question - how will search engines (esp Google) treat these pages in assessing the strength of links and therefore our page ranking. Is there any way to format the token so that search engines will ignore it and simply assess the base URL.

[Yidaki]

Google may treat the pages as doublicates ... you'd run into problems. As far as i know, there's no workaround. You could disallow crawling of all urls that contain a referer token. But this way you'd loose the PR that might get passed to your pages.

[simonv]

So how does Amazon do it without being banned.
If you search in Google for
site:amazon.com 0714837512
then you get about 18 results all with different token type additions and about 13 of those results resolve to the same detail book page...
I suspect some of them are listings put up by different associates of Amazon (and if you think about it, all those associates must have some unique identifier - yet many will point to the same Amazon page).

[sugarkane]

Many of those links aren't actually indexed by Google, as indicated by the lack of title or snippet text.

This is often the result of googlebot being blocked - it doesn't look like it in this case unless Amazon are cloaking their robots.txt ;)

More likely, either the pages that generate these links don't have sufficient PR to entice googlebot to follow (given that Amazon already have a few million pages listed), or Amazon is a 'special case' that googlebot is programmed to recognise.