TheMadScientist - 5:58 pm on Jan 8, 2013 (gmt 0) [edited by: TheMadScientist at 6:21 pm (utc) on Jan 8, 2013]
If what The Mad Scientist is saying is Google plans to change things in such a way that only a very small percentage of the traffic on Google will find its' way to our sites then that obviously sounds like an end of the world scenario for online businesses who haven't found other ways to generate business. I'm struggling to see it though. I agree the 'machine learning' aspect of Panda will become more and more influential over time and I can see how that could be the foundation of the 'one right answer' concept but my understanding is the machine learns from the behaviour of humans who visit your site, so how can it learn what the one right answer is without continuing to send humans to our sites? Has Google really developed intelligence to match that of a human? I doubt it.
I think I need more info on the 'one right answer' idea. Will this be an option, with results listed as normal if you're not satisfied with the one right answer? Does anyone know or are we discussing something based on speculation about HOW Google will implement what they've talked about?
There's actually more info on it in some of my previous posts in this thread.
Yes, they will still have to continue to send people to sites especially when they 'don't have the answer', but as things are 'segmented' and 'fragmented' and 'refined' the 'personalization' of the results will have to become much more drastic and individualized, so where it used to be you if ranked for Super Widget you would rank in the same place or close to it 'across the board'.
Once they get things 'segmented' down, rather than ranking for all 10,000 Super Widget queries you may only rank for 100 Super Widget queries if your site is determined to be 'right' for that sub-set of searchers.
I also stated earlier I'll try and explain it more in another thread when I can figure out how to do it in 'plain English', because it's really complicated to explain, but once they 'get close' (even if not all the way there), the game can really change and how the non-Google results they actually show are determined are, in my opinion, is very likely to change with it...
Can you imagine if they applied the 'image search' comparisons they make to sites to determine what colors, contrast, layout, white space and other 'visual cues' specific searchers are happiest with?
There's no reason why they can't, even if it's not implemented yet (I'm actually inclined to think that's part of what Panda does, but I don't have any 'proof', just a hunch) ... Either way, there's also no reason I can think of for them to 'not go there', because it helps them 'drill down' to 'the right answer' for a specific individual, which is what they're moving to.
ADDED: If we just stop and look at personalization of results, as it gets farther along, it's going to be increasingly difficult to 'rank for everyone who searches', which lowers traffic sent to any specific 'number one ranking' or even 'top ten ranking' across the board ... They don't have to even 'keep all the traffic' for many here to have serious issues from decreased traffic, because as the segmenting and fragmenting of results necessary for personalization gets greater, the traffic level sent to any one site for any single query drops to N% of the % of queries for the phrase searched on and the result is shown for, rather than N% of All Queries for the phrase searched on.
[edited by: TheMadScientist at 6:21 pm (utc) on Jan 8, 2013]