If You Give Google A Cookie

[AussieWebmaster]

Okay guys I wrote a piece about the various comments on the Google cookie controversy.
Check it out here (http://blog.searchenginewatch.com/blog/080628-005239) and let me know what you think.

Matt Cutts I know you have the alert on.... all sides gives best conversations

[jimbeetle]

I'm not actually too worried about Google at this moment. Right now any of Google's initiatives that might impinge privacy are like any of its products -- disparate in nature and loosely tied, if at all.

This Register article from back in May, How ComScore can track your mouse clicks (http://www.theregister.co.uk/2008/05/12/inside_comscore/), really made me sit up. I truly had no idea what types of user data could be tracked and amalgamated.

*ComScore can track *all* 'Net usage.
*it sees what you enter into forms -- even on https connections.
*it can differentiate among users on multi-user machines by their mousing and keying behaviors.
*it can use form input to match the user against -- and get additional demo information from -- credit agency and prescription benefits manager records.

Now, here's where I do get antsy. ComScore has a fairly limited amount of data compared to Google, one million stateside users, two million total. I'm not sure what the combined Google user penetration number is (toolbar, G cookies, DoubleClick cookies, and the ex-Performics now G affiliate cookies), but dare say in tech terms that it must be a bunch of folks.

Looking forward, what happens when Google puts it all together? If ComScore can do what it does today, what will Google be able to do in the future?

Lead-lined, tin-foil hats, anyone?

[B-Double-U]

I think I agree with this:
disparate in nature and loosely tied, if at all.

I am running into problems on my personal machine that aren't on vanilla boxes. I even wipe all cookies and get a different set of results than other machines without prior searches. I spoke to G about it and she immediately jumped on the chance to tell me how they were testing this out and different users may experience different result sets depending on where the search was initiated, even from the same I.P..

Maybe it doesn't matter about the cookies so much, but possibly from other sources, like google desktop and much like the discussion over here This made me giggle (http://forums.searchenginewatch.com/showthread.php?t=22976), where other behavior is being used to determine relevancy.

I'm not real worried though, because even with enough info, it gets to a point where the sheer volume of info becomes saturated and the saturation leads to an equilibrium point. Then you are back in the middle because marketing to the masses requires a "most" focal point. It simply isn't in their best interest to provide the most relevant and targeted info. If the first result was always what I wanted, I would never have a reason to click an AD.

I run our customer info through a rigorous process of SPSS batches and it seldom comes up with meaningful relationships. Sure it turns up tons of relationships.... but there just isn't a use for most of them. Maybe on a greater scale there is a need, but I think it is so far outside their scope, that I'm personally not that worried.