Webmaster Forum - View Single Post

dpam · 11-18-2003, 07:58 AM

Hey John, with all due respect, you seem a bit harsh on these guys. Obviously you don't think they really don't test the algo's internally. It seems that when they introduce them the sheer size of the data requires that results are 'mid-air' for a while while the system updates. That is understandable given the scale at which they're operating isn't it? I would also think that in the case of work like theirs, even a reasonably large internal test might not show all the inconsistencies as running it on the billions of pages in the actual data set - so reasonable testing might not expose all the problems.

This is a complicated issue, and the more I read about Google in general in SEO forums or newsletters, the more I see the debate reduced to simplistic and often immature posturing (just like we get on CrossFire or Face the Nation :-) ) These guys are in a tough spot; a couple of college kids set out to build a reasonable list of web sites and apparently succeeded beyond everyone at the time. I'm fairly certain they did not intend to become the arbiters of who goes where in what quantities for the entire internet, yet that is what happened. So their responsibility went from producing a 'pretty good' list for surfers to being the primary economic driver for tens of thousands of businesses - the expected tolerances for those two activities are orders of magnitude apart. Surfers accept just about anything, businesses expect near perfection. So while they're clearly benefiting from these unintended consequences (adwords) they're also drowning because of it (the 'SEO' gaming of their results).

So now they have to live up to it, or they'll die. Just like anyone who hits it big young and then has trouble repeating or maintaining their success. Dylan once referred to it as having to 'do intentionally what you once did instinctively'. That is very hard - which is why most bands suck by their 3rd or 4th album (not to mention the one-hit-wonders) and it's one reason why many dominating businesses get beat anyway. (not trying to lecture, just on a roll here...).

My point is that the challenge Google faces now to maintain their lead is much more difficult than it was to take over the industry from their dorm room. Finding an algorithm that cannot be gamed, or can only be gamed with positive results is a damn tough gig when you're up against the collective brainpower of nearly everyone in the world. Plus, they're going to start drowning in these 'keyword' lawsuits and removal request as trademark holders figure out the real power of this system and they don't want to be gamed either. (I think this horrible trend is what will really kill this business, but that's another rant.) Lastly, the rampant fraud that is almost certainly pervading the adsense system will come to light, and one of their financial legs may tumble.

If I'm them, I'd IPO damn fast.

So this message started out in their defense, and I do think they make an honest try of it although there are undoubtedly some politics in it we all don't know about and some conflicting priorities that mean 'relevancy' doesn't win every debate. They may stuble and even have to back up sometimes but I'd rather see this than a screwed up result that isn't getting fixed - that only serves to escalate the 'black hat' arms race to the point where all results will be useless to everybody. It seems unfair to criticize anyone for an honest attempt to try and improve.

But along the way of making that point it became clear to me that they peaked six months or so ago, and the weight of this success is starting to drown them. Smart people figure out how to come back, so we'll see if they do or not (although I think the emergance of Yahoo/Inktomi and MSN Search within the next year is almost certainly going to take them down a few notches not matter what they do.)

For whatever that is worth....

11-18-2003, 07:58 AM	#77 (permalink)
dpam Inactive Join Date: 10-13-03 Location: Nearly the East Coast Posts: 35 iTrader: 0 / 0% Latest Blog: None	Hey John, with all due respect, you seem a bit harsh on these guys. Obviously you don't think they really don't test the algo's internally. It seems that when they introduce them the sheer size of the data requires that results are 'mid-air' for a while while the system updates. That is understandable given the scale at which they're operating isn't it? I would also think that in the case of work like theirs, even a reasonably large internal test might not show all the inconsistencies as running it on the billions of pages in the actual data set - so reasonable testing might not expose all the problems. This is a complicated issue, and the more I read about Google in general in SEO forums or newsletters, the more I see the debate reduced to simplistic and often immature posturing (just like we get on CrossFire or Face the Nation :-) ) These guys are in a tough spot; a couple of college kids set out to build a reasonable list of web sites and apparently succeeded beyond everyone at the time. I'm fairly certain they did not intend to become the arbiters of who goes where in what quantities for the entire internet, yet that is what happened. So their responsibility went from producing a 'pretty good' list for surfers to being the primary economic driver for tens of thousands of businesses - the expected tolerances for those two activities are orders of magnitude apart. Surfers accept just about anything, businesses expect near perfection. So while they're clearly benefiting from these unintended consequences (adwords) they're also drowning because of it (the 'SEO' gaming of their results). So now they have to live up to it, or they'll die. Just like anyone who hits it big young and then has trouble repeating or maintaining their success. Dylan once referred to it as having to 'do intentionally what you once did instinctively'. That is very hard - which is why most bands suck by their 3rd or 4th album (not to mention the one-hit-wonders) and it's one reason why many dominating businesses get beat anyway. (not trying to lecture, just on a roll here...). My point is that the challenge Google faces now to maintain their lead is much more difficult than it was to take over the industry from their dorm room. Finding an algorithm that cannot be gamed, or can only be gamed with positive results is a damn tough gig when you're up against the collective brainpower of nearly everyone in the world. Plus, they're going to start drowning in these 'keyword' lawsuits and removal request as trademark holders figure out the real power of this system and they don't want to be gamed either. (I think this horrible trend is what will really kill this business, but that's another rant.) Lastly, the rampant fraud that is almost certainly pervading the adsense system will come to light, and one of their financial legs may tumble. If I'm them, I'd IPO damn fast. So this message started out in their defense, and I do think they make an honest try of it although there are undoubtedly some politics in it we all don't know about and some conflicting priorities that mean 'relevancy' doesn't win every debate. They may stuble and even have to back up sometimes but I'd rather see this than a screwed up result that isn't getting fixed - that only serves to escalate the 'black hat' arms race to the point where all results will be useless to everybody. It seems unfair to criticize anyone for an honest attempt to try and improve. But along the way of making that point it became clear to me that they peaked six months or so ago, and the weight of this success is starting to drown them. Smart people figure out how to come back, so we'll see if they do or not (although I think the emergance of Yahoo/Inktomi and MSN Search within the next year is almost certainly going to take them down a few notches not matter what they do.) For whatever that is worth....