Miamacs - 8:21 pm on Nov 10, 2007 (gmt 0)

I see more myths created in the OP than busted but that's allright, the less competition the better.

...

As for how deep they are, you only need to pair up bounce rates (ie. 1 p/v, exit ) with time spent on site to get an idea on whether that was a visit or a(n) ________ . ( insert whatever you feel fit )

...

As for themes, topics, PR, I'm comfortable with the top 10 in a month results the way I build links, as for their numbers, they are rarely more than a couple of dozens. Topics and themes are mentioned btw in a categorization page at Google, although not affecting organic search directly, there's a pretty good list for categorizing sites that are currently running AdSense. Mind you it's official.

But again, you may carry on telling all these things, for it's pretty accurate, if only for low-profile, low-competition, unpopular or just way too special searches that the 'dumb computer' has no idea about.

For you see, the dumb computer is actually but people ( linguists among them ) of whom some look for patterns, while others write programs to gather data, then yet other programs look for signals that match the patterns.

Meaning if your area of expertise in Google is still in 2002 mode ( high PR links, fast, nothing else matters, there can be only one etc. ), it's probably because there isn't enough data on it. No trust is calculated. No relevance is known to major or mainstream themes and topics. No phrase sets to check co-occurrance for no stop words, for what it's worth it could be written in martian... basically, it's a free for all sector.

I see this kind of enthusiasm all the time from people marketing unmarketed stuff, or who found a loophole in the phrase sets Google currently ( algorithmically ) monitors, only to see how meaningless this approach is in areas where a single, well defined, ultra competitive keyword can be used to build an entire company around it.

I have websites in areas where, as you mention, it's about PR, there's no link profile check, and no topics set at all. But that doesn't mean this is the norm and I can dismiss the knowledge I gathered throughout the years, rather that - as I've predicted for these sites too - it's in a sector yet to be explored by most of Google's modern age algos and filters. Either because it's not written in a language Google can translate ( hint hint ) or because it's not popular enough. But I have mainstream sites to care for so... let me believe in some of the stuff you dismissed as myths just a little bit, until I can come up with news ones.

...