Slashdot Mirror
NYT On Google's Role In Internet Advertising
prostoalex writes "John Markoff and G. Pascal Zachary from The New York Times take a look at Google, its already dominant position in the field of Web search and its increasing influence in the field of Internet advertising. Google is driving advertisers away from larger advertising venues, like AOL-TW et al., since (surprise!) people actually pay attention to relevant text links and are quite annoyed by pop-ups and similar "innovations". Some interesting data about Google: number of employees is about 800, number of buildings is 4, number of servers is 54K, for which there are about 100K microprocessors and 261K hard drives. This is claimed to be the largest computing system in the world, and that also raises barriers for anyone entering the field of Web search - most of companies out there can only imagine a Beowulf cluster of these, let alone build them so that the Web searches are delivered within a second."
I dont know why its so hard for you guys to create an account, and have your browser just autofill it each time you visit. I mean, if you use it daily, or at least weekly from a slashdot article, why not use a real account?
I mean come on, they let google index articles off news.google.com.
people really do click those ads! I've always tuned them out, so I wondered if they could really make money with them. Apparently they've been quite profitable. I hope they keep things the way they are.
Well, I don't know about you, but I'm not all that fond of them filtering results based on where people are searching from.
And they take all manner of porn ads but the only alcohol related ads are for hangover cures - so exploiting messed up men and women is ok, but exploiting your own liver is not?
lets see, my last 2 real accounts got mysterious deleted (maybe too long between visits?), and maybe I just don't want to have my privacy invaded.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
Does anyone one else think Slashdot should have a Google topic icon?
Or maybe just an NYT icon....
But seriously, so many have been added lately, what's one more going to hurt, especially since it's a company/search-engine discussed here so often?
Um, let's see $1,000,000 / 54,000 = $18.50 Hmm, for dual cpu, 4GB ram, rack mount machines I think you will need a bit more than that =)
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
Very true :) There seem to be no limits in life as to what men will turn into a silly 'dick-size' contest. Men turn practically anything and everything into a contest, no matter how pointless and trivial something is (or for that matter, no matter how lofty and meaningful something is). I think its one of humankinds more useful traits, competition 'gets the blood going' (figuratively I mean) and spurs us on to be better and to make better things.
Don't let me be the only one here who has not only used an advertising link from google, but has actually bought something from one.
I find that thse links off to the side actually aren't annoying. They are off to the side. They dont interfear with my search when I'm not looking to spend some money, but when I do search for something to buy they usully come in handy. At the very least, it indicates that the store has some income with which to advertise and is not being run by monkeys. Just my $.02
My UID is prime and so is this number: 09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0.
You'd be amazed how many windows machines one person can administer if every time a machine went down for a hardware or software fault you didn't even bother to take it down from the rack- just removed it from the list of good machines. For google it just doesn't matter how easy/hard it is to administer the machine. They image it once, and never touch it again.
The second pro-Google argument is that doesn't reorder results in a naughty way. Yet countless articles provide evidence of results changing from moment to moment, pages appearing and disappearing. In some cases, censorship of sites has occurred.
Encompassing the above argument is that Google doesn't put sponsors higher. Since almost all the big companies employ "search engine optimization", the only difference here is the technique a company must use to get listed higher. The small but relevant web sites who don't spend time "marketing" with good keywords / spamming with duplicate sites still find themselves lower down.
The most irritating pro-Google argument is that it is innovative. No it isn't. It merely uses techniques that were being developed in the early 90s but omits all the later intrusions. Google is not a technology company, but it has excellent marketing staff who realise that "intrusive marketing" doesn't work. That's all.
I can't believe they would leave thousands of dead machines powered up and taking up rack space. Surely it would be more economic to hire anyone to swap them out or at least remove them and save paying x-thousands per week hosting costs... In any case, presumably someone originally made the descision to buy all these machines, so they would be replaced anyway eventually.
Code, Hardware, stuff like that.
In a beowulf cluster the memory isn't shared, well unless you use one of those new shared memory things that they have (sorry forget the name) but either way its usually done over Ethernet or Token Ring. Most things which are memory intensive aren't good candidates for beowulf clusters because the movement of the data in memory takes so long. Most beowulf clusters use PVM and MPI to pass messages amongest themselves and that's how they communicate. The article doesn't give too much data on Googles software, but if their server application passes data amongest all the servers then it could be considered a beowulf cluster.
With regards to the SGI Origin aspect, its probably pretty easy for them to just script everything using linux and they can get everything pretty automated. Plus they can easily add boxes to the datacenters that need them, so that load is more balanced. With using a single system image, you probably couldn't just let the computers go like google does (just leaves the dead computers dead). Although the fact that google's database could then be stored in one or two computers might be helpful.
There isn't any way that the NY Times is going to be slashdotted. Mod this parent down, what a waste of a post and of the flow of the conversation. As has been pointed out countless times there is a ring around their subscription service and even that in my opinion should now be modded down as repetitive.
Of course they'll die if people like you force them into removing ads--because that's their income. How do you expect them to exist without some form of income?
--RJ
All the stuff you mention is legally mandated by some country or another. Porn ads online in the US are not generally regulated. Alcohol ads anywhere are closely regulated. Note that you won't see a porn ad unless you are searching on porn-related terms.
In some venues, search results are regulated. If Google pulls something from the German site, (for example) it's often because they got a nastygram from a German court. Note that it's still available to searchers in other countries.
So unless the proposal is that Google should go lawless and take on all the world's governments at once, their behavior is pretty understandable. Hey, I don't like driving below 65 either, but I do it! (At least when the cops are watching.)
But I am an advertiser at google. People here seems to be complaining about the ads at the side.. etc.etc.
Looking at advertising at the top pay-per-clicks (Overture, Google, Findwhat), Google is the only one that has Instant Gratification. We created an e-commerce site and were able to start driving qualified traffic to it in about 15 minutes. With Overture or Findwhat, we would have had to wait for several days to a week, and to top it off, they might have rejected many of our listings through their brain dead editorial process.
Google at least is fair in the way how they reject listings... they have editorial guidelines, so you know upfront, and secondly... if your listings suck for relevance they get automatically booted.
I think the NYT writer fails to take into account the instant gratification factor, which IMHO is the greatest advantage to using Google. Because you can test your business model right away. If it sucks, then you can take your business model offline before it gets too late.
Finally, because of Googles contextual ads (some of which are shown on Slashdot), they have really co-opted advertising on the web. Because of this, Overture's stock is in the barrel and I think they will become a no-player in the near future, simply because of pending moves by Yahoo and MSN, their largest search suppliers. Even though they've bought Altavista and Alltheweb, when was the last time you saw traffic coming in from those searsh engines into your Apache server logs?
My only fear with this, is that Google can become too powerful (see Microsoft), and can then call the shots with advertising on the web in general. We saw that behavior with Overture, just before Google launched their program.
Don't laugh, we revel in Google's friendliness, relevance, and geek cred right now, but I hope they don't go public too soon. I hope that Page, Brin, and Schmidt hold on to the reins tightly for the time being... because once Wall Street steps in, the ride for the consumer is over guys!!
Newsfollow.com
Removing the "backdoor" gains them NOTHING.
/.ers read the article through the archive link. If they get rid of this link, the same article would precipitate:
Hopefully, they realize that those that "incorrectly" use the backdoor wouldn't give them valid registration information anyway.
Let's say 10,000
- 5,000 "nyt_suxors_2003april" with password of "password"
- 3,000 "initial_lastname_04_2003" with a password of "password"
- 100 people that register correctly
and the rest not reading the article (or waiting for a karma whore to post it).
Frankly, that's a lot of bad data that can be avoided by providing the archive link.
- Tony
Hell, 50% of their searches are probably FOR porn, why would they filter it out? And as for alcoholic filters, they are a private company, they may do as they wish.
$10m for 54000 servers? Considering these servers are said to average 2CPUs and five hard drives per, I'd say your estimated $185.00 per server is a little on the slim side.
Estimating a more round $2000 per server, we come to a figure of $108 million. Factor on top of this the costs of housing these servers, including backup power (UPS and generator), the real estate (you can't shove 54k servers into a spare equipment closet), the custom software (which is where the real money's at) and the costs are suddenly a lot more real.
A project the magnitude of Google isn't something that can be implemented by a company as a side effort, which is precisely why nobody, including Microsoft, Yahoo!, or others have had any luck thus far in duplicating their success.
BD Phone Home!
Shameless plug. Like you weren't expecting it.
Follow this link. The title of the page is "10 things Google has found to be true".
Look at number 6 it's titles "You can make money without doing evil."
Maybe that's a strong reason why people continue to use google despite the competition that keeps popping up (hey whatever happened to snap.com?).
I admire any company which holds as one it's core values a commitment to not doing evil. Unfortunately they are in the minority.
War is necrophilia.
Time to go metamod some moderators i think..
If Google really cared they would fix Android Chrome to reflow text, instead of discriminating