Silly Rabbits: Google is for Spam not for Search
smoothspan.wordpress.com
smoothspan.wordpress.com
That's already happened. I am pretty sure that this is not intentional though.
What would sway me is if say 'bing' would have zero spam and google would have a whole pile of it.
Is there any research confirming or denying that?
AFAIK BOSS is still Yahoo technology, not bing. There may be a plan to migrate BOSS to bing.
Has anyone from inside Google said anything about this yet?
> Has anyone from inside Google said anything about this yet?
Matt Cutts and moultano are here all the time, saying that they're working on a solution. But for some reason, it is better to make up shit and then believe it than to actually listen to the people who talk.
People don't say "follow the money" in TV shows for no reason.
What you don't realize is that since Google went IPO, it's not all about building an amazing search engine. It's a lot about revenues and net profits. There are lots of businessmen in charge who want to make lots of money quickly, and couldn't care less about the quality of the search results.
First of all, Google went public in 2004. I would like you to point me to a study showing the steady decline in quality since that time. Who are these 'lots of businessmen' you talk about? One of the complaints about Google has actually been that they prefer to have engineers in all sorts of roles including management.
Now, since you've gone all conspiracy theorist, I have a theory of my own, I hope you will excuse it. My theory is that the pro-Apple blogging community has gone ape-shit over a perceived threat to iOS from Google, and are now looking at them as 'The enemy', and this whole rash of blog posts maligning Google is part of a larger 'Death by a thousand paper cuts' strategy to make Google seem uncool, and therefore, as we all concur, dead. I also believe that said bloggers are in fact undercover employees of Apple who get paid by the number of times they mention Apple products in reviews of their competitors' products, but that's just me being nuts.
And then you call me a conspiracy theorist, immediately followed by a pretty insane conspiracy theory about Apple community. I'm not even an Apple fan.
What I meant to say was that I've seen many reports which show a large percentage of executives at Google to be engineers and programmers, although I can't find a link right now. However, I seriously doubt that Google engineers would have gone along without revolt down a slippery slope as the one implied in the original article.
The facts are
1) Google search results are full of spam.
2) Many of the spam sites have been reported (I reported a few myself) and blogged/written about for years.
What can one conclude from that? I think there are 2 possible explanations
1) It's not in Google's financial interests
2) Google is failing at filtering spam
I doubt it's true, because simply banning the top 1000 spammy domains would cut more than 99% of the search spam. It's something that an intern can do.
If Google wants a pure algorithmic solution instead, well, good luck to them. I think until AI is smarter than humans, such solution doesn't exist. Spammers can always find loopholes in algorithms.
It most likely a much simpler explanation: web results with content explicitly generated for a specific target search string is simply getting closer and closer to being actual genuine content, and thus harder and harder to differentiate and filter.
2) Google makes a ton of money from spam sites.
• As a matter of policy, Google doesn't want to remove borderline spam sites (content mills, "$keyword free download" aggregators). Like Experts Exchange they might be spared as long as they bend, but not break the rules.
• Google really doesn't want to start mass manual banning (because that might be manual banning vs automated generator), and some spammers found way to sneak past Google's dupe detection algorithm.
Google stands to make the most money in the long run by being the preferred search engine for the most people. Right now, they are. But that could change if another search engine can convince people they deliver results with less spam.
Not necessarily -- if other search engines drive people to pages full of Google ads, Google still get paid, right?
Is it true that google has more spam than other search engines?
That's a conflict of interest, one well spelled-out in the original post: When removing all spam (let's pretend it's possible) will negatively impact revenue, even short-term, there's less motivation to do it.
Wouldn't the people who went to spamy sites go to the "good" sites and still click on these sites thus still make google money?
Unless you are suggesting that the spamy sites are the only ones available and if they disappeared searchers would not be shown results, I do not quite see why removing spam would hit the bottom line.
(1) If the natural search result landing pages are slightly lower quality – awful writing created via lowest-bidder/freelancer content mills – but the AdWords landing pages remain higher-quality, due to more investment and ability-topay filtering, then AdWords results become more attractive to users. That's more clickthroughs and more revenues.
(2) If the first page of results reviewed by a user doesn't resolve their intent, so they reformulate a more specific query, Google gets to display a second, perhaps better-targeted set of ads on the second query – a second chance at satisfying the user goal via a paid placement rather than an unpaid result.
(3) If there are good natural results sites for a topic that lack ads, but others can be created with roughly the same info that have AdSense, unless Google specifically punishes sites for containing AdSense, the relative number of AdSense sites with duplicate info will multiply over time. Even if they are ever-so-lightly better (or 'as good' to low-literacy readers), and thus appear 'good' in some of Google's algorithms, they wind up wasting time in aggregate with duplication, and dilute the traffic/community of original ad-free source sites. More likely, though, they are just-enough-better in the cynical 'SEO' dimensions to display other less profit-maximized sites, and thus may be worse in the (not directly measurable) value-to-readers dimensions. Still, in the short- and medium-term, the multiplication of such sites, and replacement of AdSense-free sites with Adsence-drenched sites, makes Google money.
11159 karma points does not give you the card to play mr arrogant gojo.
But thank you for your explanation.
I think the first point is quite clever and does make sense, but it has to be balanced with the number of visitors that google may loose by allowing the index to become spamy.
The second point seems to be what everyone is thinking of and perhaps the strongest argument to suggest that google might not want to tackle spam. For that argument to gain any credibility however it needs to be shown that other search engines, i.e. bing, which perhaps do not have a conflict of interest are doing better at showing quality results. Otherwise, I think it makes it more likely that there are other reasons than conflict of interest.
As to the third point, the cases that it would apply to I would think are miniscule in the grand scheme of search, not least because a very tiny minority of websites does not use ads, excluding e-commerce.
Finally, I think a strong argument for google to tackle spam, and against the proposition that google does not want to, is that google would make much more money by converting more clicking visitors into customers for the adword landing page. If spamy sites have a much lower conversion of clickers for the adword customers, then cutting off these spammers would increase the conversion rate, leading to higher payments for clicks to google by adword users.
Pretending these low-literacy users — 43% of the US adult population! — don't exist, or can't be discussed frankly, is less respectful than seeking to understand them.
It's quite likely that these people are late adopters of web search, and thus their proportion of Google users is still rising. Consider, also, that there is evidence that lower-income, lower-education users are more likely to click on ads.
I've been perplexed when discussing content mills with Googlers how they quickly assert (often in similar phrasing, suggesting a party line passed along in internal or official communications) that reasonable arguments (or very subtly, internal studies) imply many users appreciate the content mills' writing.
Perhaps it's the low-literacy users who are happy to land on a mill site. Either they can't tell the difference from a quality, authoritative treatment — or maybe even the plodding, keyword-stuffed, repetitive writeups are actually reassuring for slower readers.
There has to be a logical reason, other than Google venality, that such awful spamglish writing decked in AdSense ranks so highly.
But thanks for the link, very informative. I suppose if 40% of internet users are low-literacy then such sites as about.com sort of make sense.
I think we need to differentiate between content mills and spam sites though. Although many do not like content mills, I would not consider them spam but low quality content.
Spam sites seem to be more reliant on affiliate marketing, especially in the review of products niche, as well as monetised by adsense.
I do however think that it would help the discussion immensely if adword users can provide some data as to how does the quality of the site compare with conversion. I would assume that the higher the quality of the site content, the better the conversion, but, assumptions can be wrong.
In the long run, it's a losing prospect for Google: short term profits at the cost of long term reputation and trust doesn't make sense for a company as large and as public as they are.
Edit: I know they experimented with voting arrows on the results page and I'm curious why some variant of that hasn't been deployed more widely.
No idea how seriously these reports are taken, and whether or not something like the fake Mastercam blog would be considered spam.
I'm not going to say that Demand Media content is world class, or even good, but it's a far cry from scraper-spam and I don't think it's honest to group the two.
Even in an automated scraper scenario, things aren't clear cut. How much processing needs to be done on source data before it goes from illegitimate copying to value-add? Fundamentally Google is just a really really amazing scraper system.
Same thing with content farms. At what point does original content go from "paraphrasing Wikipedia" to providing useful summary information? Should Wikipedia now hold a defacto monopoly on summary-style encyclopedic information?
Scraped content and empty directory pages (that washing machine reviews page that has nothing more than an H1 and some crummy auto-generated sentences) vs... Mediocre human written content from sites like e-How and other Demand media properties.
The first kind of content is pretty clearly spam, and doesn't deserve a place in the index. Everyone can agree on that, and I'm perplexed as to why Google hasn't taken more action against these types of sites. I have a hard time believing they can't do it algorithmically.
The second type isn't as clear cut. Most of the readers here and on the cited tech/search blogs know what this content is and where it comes from, and are probably a bit more critical of it than the average user for that very reason. I agree that these article directory pages are usually painfully mediocre, but I wonder if they aren't "good enough" from the point of view of someone who doesn't know better. After all, I havent seen much of an outcry about search quality coming from the mainstream; it seems limited to HN and other tech circles.
So what's Google's benefit to annoy everybody with non-quality search results? I'd guess, the SEO is currently just too good.
I agree with your hunch; however, bounce rate is a relatively bad thing. If, using analytics, you see that most users navigate away from your page within 30 seconds, you probably don't have engaging content. How is that different than the user clicking on an AdSense ad because the content on a spam site isn't connecting emotionally or intellectually with the users? If anything, Google should be giving a negative score to sites that have a huge percentage of their users that actually click on ads to continue the search for what they were originally looking for... though that is basically the opposite of their business model, to have users click on ads.