|Wall Street Journal Points Out Google Spam|
Search for "god" pulls up MP3, PHP-Nuke in top 2 Spots
Monday's Wall Street Journal has a story questioning whether Google is maintaining its lead in technology, using a couple of sites that rose to the top inappropriately as examples.
The story points out how sites are manipulating the results, as evidenced by the first two results for "god" - an MP3 site and PHP-Nuke.
The story points out the related searches features of Teoma and Alta Vista and notes that Google has no similar capability...interesting reading.
These sorts of "off the wall" results are proof in the end that you cannot have 100% automated SERPs and get 100% good results.
I have often said that google needs to have a team put together to check the top 20,000 or so seach terms each month to assure that the top 10 sites in these SERPs are quality sites, and that they are relevant to the search term. This would improve the end user experience for a large percentage of searches, and would make it easy for Google to self police and spot spam with human eyes, rather than relying only on spam reports and disgruntled end users.
Let's see... review 10 sites per serp ( at least looking at the page as listed) takes maybe 10 minutes per serp. that would be 200,000 minutes of work per month. call it 3000 hours, round number. Basically, 100 hours per day, so maybe 15 people. Maybe 50k per month of expenses to massively improve the results that most people see, and to block out spammers from the most popular results, making spamming google a much less profitable process.
Alas, Google is all about being algo driven to as close to 100% as possible, so I don't think we will see this really happen (and if it is already happening, the people scanning need to learn how to spot spam better, because too much is getting through!).
Not very new stuff [webmasterworld.com]. If a lot of people put "God" as the descriptive text in a link for MP3 and PHPNuke it will affect the position of a keyword in G. Old news and really bad story with no recent data by the WSJ!
I wonder if they got the idea for that search from here:
#1on God [webmasterworld.com]
PHP Nuke no longer #1 for God [webmasterworld.com]
|Old news and really bad story |
Not sure how spammy search results that come up today are old news - but sorry you didn't find the fact that Google spam problems have made it to the pages of one of the country's most respected and widely-read publications interesting :)
Amusingly, in the source code of the page of the site that displaced PHPNuke for "god", you can see that they are also linking to pornographic sites: "busty....com" - I wonder if G is going to penalize "God" for being in a "bad neighborhood"? ;)
>>These sorts of "off the wall" results are proof in the end that you cannot have 100% automated SERPs and get 100% good results. <<
On the other hand you cannot have human reviewed systems and get 100% good results e.g. Dmoz, Y! directory.
I dont think anybody expects to get 100% good results for any search, especially if they are so silly to do a search for "god". I mean, heh, what practical motivation would there be for doing such a search except for keyword research, to catch a search engine out, creating a peice for a humour/satirical mag, or to spin a story on a low news day. I can see a justification for "names of gods", and many other queries but searching for "god" is about as practically sensible as doing a search for "talentless hack".
Basically if anybody, including hacks at the WSJ, want to do a search for a generic one-word term any day of the week in any search engine, they will get what they deserve.
|Basically if anybody, including hacks at the WSJ, want to do a search for a generic one-word term any day of the week in any search engine, they will get what they deserve. |
Accurate, on-topic results? :)
Hey! All the SEOC can now spam the church and tell them they too can all be number 1 on Google :o)
On a serious note though (the above was a joke SEO's) the number 1 spot is a redirect to s site where God is nowehere to be seen.
I have never seen any possible benefit in abusing keywords like "god" "sex" etc if the site is totally unrelated?
Quite frankly, for the WSJ to plundge to such hair splitting depths wont win any confindence from anyone. Let's face it, what on earth does the WSJ and God have in common.
>>what on earth does the WSJ and God have in common?>>
off topic but for information "God" is the default admin username in early (maybe still) versions of phpnuke.
Mardi Gras is correct that the WSJ story is significant. Not because it is a good piece of journalism, for it is not: it ignores the better quality crawler processes of Teoma and AlltheWeb, for example, and is based on a truly trivial search example. And it does not (to my satisfaction) argue the case against Google’s popularity-based algorithm, with its devastating effects on information choice and scope.
But the WSJ piece is significant because it shows *that the mainline business and investor community is beginning to take notice of us.*
Search quality is vital to all of us. Some of us are improving it with next-generation search tools, where the quality issue is being resolved in ways that Google simply cannot match. Others are working hard to try and ensure that the spider-based tools at least tend to rank sites more accurately. The entire paid-search portion of the business, from Overture and Google down through the hundreds of copy-cat PPCs and thousands of false-front affiliate “directories” is in a state of rapid change, as advertisers and website owners begin to learn that the competitive keyword bidding business model can be a disastrous waste of money, not to mention the time required to maintain bids.
But as all-consuming as these trends and issue are to you and me, they mean zippo to the average person, even the relatively educated and affluent ones reading the WSJ.
Any evidence that the general business audience is waking up to the realities and scope of Search Engine Marketing, even in the form of that weakly-written WSJ story, is highly important, since it means our crazy, still highly marginal industry, is on the verge of becoming mainstream, with all that that implies.
A short side-bar comment:
Raw Alex is spot-on with his observation about the flaw in any 100% automated indexing and ranking method. To say that human-reviewed techniques don’t work, and to cite DMOZ as the evidence is a weak arguement, since the OPD is effectively un-managed, and since it has no economic model that could make it effective as a service.
LOL! Perhaps the journalist should have used this keyword rather than "God", the results are more relavent. I think WSJ is result number 124,000 from 125,000
>>To say that human-reviewed techniques don’t work, and to cite DMOZ as the evidence is a weak arguement,<<
Nope. not weak at all. I didnt say human reviewed directories etc dont work, and the two biggest examples we have are Y! and dmoz are fair game as examples of leaders in the field. Neither of them are disasters, but they are hardly better at finding relevant content than robot engines The reasons are cost to pay humans, (meaning much reduced reviewed databases) and human subjectivity which can never be removed. 100% spidered indexes also have their own different problems, mainly in countering spam. But as far as i know Google rankings are already influenced by human review in the incorportion of the dmoz directory and hand review of problematic sites.
I also agree in the significance of the story for the reasons you and Mardi Gras say, but likewise taking that article as an argument why "machine-generated" indexes dont work seems flimsy and opportunistic. We are SEOs and webmasters who knew these issues already; the WSJ target market are 99% influential "leaders" of industry. But i still wonder that to an audience of these types why they want to use a theoretical example like "God". "Economy", "Business Management", "global trade", "creative accounting", "enron" and "corporate jets" would be far more sensible to that lot, for which the results are, of course, well on target...
Google has aspirations of becoming a public company right? The WSJ is a paper well-read by the investing community. They as well as others can help make or break a company's stock price. Some of their articles might be slanted, but that's the way the world goes.
Google will either take notice and do something or they'll continue to be picked on which will impact their abiilty to raise funds in the marketplace.
|The WSJ is a paper well-read by the investing community. |
Good. More stock for me to buy. :)
Anyhow - a search engine is meant to help people find information - typing in "god","jesus" or "music" is not a very clever way to find information... They're just being journalists...
The significance of this article is not in its content per se, but rather that the WSJ made the editorial and management choice to run a critical article. Maybe they're just trying to build up a history of looking at both sides. That would be my guess.
Plus, any leader in any industry is going to be "picked on" - that seems to be one part of human nature. We root for the underdog, and after they succeed, we attack them. What a game we play!
The #1 site for "God" isn't about Eric Clapton. Thus, I say this is a bad SERP.
My feeling is that if you are looking for information on something as broad as god -- which could mean Christian, Jewish, Druidic, Pantheistic, Roman, Greek, Muslim, African, Asian, rock-n-roll, baseball, soccer, etc. -- then they should be using more than one word to do a search.
This is not spammy results, but poor search technique.