Go Back    Forum > Digital Life > Tech Myths, Misinformation

Reply
 
LinkBack Thread Tools
  #1  
10-02-2011, 10:47 AM
kpmedia's Avatar
kpmedia kpmedia is offline
Site Staff | Web Hosting, Photo
 
Join Date: Feb 2004
Posts: 4,311
Thanked 374 Times in 341 Posts
One of my favorite web technology myths is the one about SEO hosting (the attempt to fool search engines into giving your sites more rank). Like most myths, it has some basis in historical facts, but it's been distorted through the years. Spamdexers (a.k.a. most so-called "affiliate marketers"*) love to spread this myth, and several sleazy hosting companies will gladly take your money for something that's proven not to work.

This post is answering a question that was asked here: http://www.digitalfaq.com/forum/web-...-best-web.html
However, it was made into a new post, so as not to sidetrack that discussion on high quality web hosts.


Quote:
So I am very keen to read your understanding about SEO hosting. I used to think that having your sites on multiple IPs was a plus, with links out from these sites to my clients' sites all being counted by Google to help the clients' sites all climb up in Google.
The idea behind "SEO hosting" is that you can fool a search engine into thinking related sites are unrelated by putting them on separate IP addresses. Then because the sites are supposedly unrelated, links between them will raise the value of the site to a search engine, resulting in it being displayed in top positions in the SERPs (search engine results pages -- the pages that appear after something is searched for).

It's currently 2011. Once upon a time, some 5+ years ago, that did work to an extent. But now? No. The search engines (mainly Google and MSN-now-Bing) caught onto this long ago, and relationships are now figured with other (mostly obvious) vectors. If anything, trying to trick the SEs will more likely land you in a sandbox, as opposed to simply being devalued as a link scheme.

From 2006-2009, Matt Cutts (the public face of Google) was writing articles and making videos, discussing how the technique was identified and foiled by the search giant**. He explicitly states, in no uncertain terms, that this will not work. The common defense by SEO hosting proponents is that Cutts is lying, accused of some nefarious purpose that varies from wingnut to wingnut. It's conspiracy nonsense.

Advances in virtual hosting (shared hosting) have made it possible to have thousands of websites hosted on a single IP address. Servers of today are on par with the super-computers of the 1990s. A single 4U server with redundant Intel Xeon CPUs (using several dozen cores of power) and dozens of gigabytes of RAM can easily maintain vast numbers of websites -- especially when using webservers like nginx or LiteSpeed, and an OS like CloudLinux. It would be insane and asinine for a search engine to penalize modern technology. Think of all the blogs hosted with HostGator. Given the "community" nature that blogging can have, linking to one another would be a natural process of bloggers, and Google/Bing understand this.

Also...

Link quantity no longer plays a primary role in search placements anyway. So the entire process is futile. Search engines want to present users with relevant, authoritative, reliable, non-spammy content***. You can have more links than a #1 spot for any keyword search term, but still find yourself on page 2 of the SERPs because your "content" has been deemed unimportant or outright crap. So focusing on link schemes is a fruitless endeavor, especially long-term.

The same is true of PageRank. A higher PR site can be further down pg1, or on a later page entirely.

Google has become a bit more obtuse on the specifics of what creates top rankings, but for most results the SE simply returns a quality site that answers a search query -- not just any site with high inbound links and high PR (that also answers the same search query). A top result could have only modest inbound links, and a modest PR, yet still be deemed worthy of that top rank vs other higher-ranked (PR + in-bound %) sites/pages.

Quote:
And that seems to be working in my experience.
In all likelihood, you're having a placebo effect. It's applied false logic. You've tried to cheat with multiple IP addresses, and you've noticed your ranks rising. However, it's unlikely that this was the catalyst for your new rankings. The content of the site, quality of inbound links, and other factors are the more probable real reasons for improved rankings. In some cases, it's not that you've gained rank, but rather that somebody else (or several somebodies) have lost theirs. Their site(s) could be getting stale/abandoned, or maybe it was devalued for being caught for implementing black-hat strategies.

Even if the IP trick is somehow working for you, it's always temporary. I've dealt with many people in the past 2-3 years who want to prove me wrong. As soon as their site hits a page 1 spot, they gloat. But a few weeks later, the result is gone -- sometimes buried on page 5 or further. Maybe missing entirely (sandboxed). And as you would surmise, follow-up questions are ignored by the gloating website owner.

Quote:
But while doing the research that led me here, I found several well expressed views that suggested this is all wrong... and that Google knows all the IPs that belong to any multi-IP web hosting service, so they say dont bother using "seo hosting". They see it as a marketing ploy by hosting firms to sell more.
You very likely came across some of my forum posts in various places online, as well as posts/blogs by marketing professionals and hosts. And you'll generally notice the dissenting opinions (as blog comments or forum posts) always come from hosts that sell SEO hosting. And to a lesser extent, from obviously gullible amateurs who have worthless spammy sites in the signatures -- a.k.a. customers of such services who are afraid of getting their worldview shattered, and therefore feel compelled to defend their purchase (and facts be damned).

There are things in this world that are sold simply because people are stupid enough to buy them. Sticky labels for DVDs, for example. It's a well-known fact that a DVD can easily be warped or thrown off-balance with a sticky label, thereby making a disc unreadable in a computer or DVD player. However, there is a market demand for such labels (by clueless consumers), and therefore such labels and manufactured and sold. The local gas station sells caffeinated toothpicks. I was standing in line one day, when the guy behind me saw these on the rack by the register. His comment was hilarious: "What person would have a big meal, and then say to himself 'You know, I could really go for a caffeinated toothpick right about now.'"

For a hosting company, this is a goldmine. They can sell you a block of preset IP addresses (as shared with other customers -- the IPs are not dedicated), and then charge you a high fee. For them, it's pure profit. If you're too dumb to not know any better, that's not their problem. In fact, the more shady/unethical hosts go so far as to defend SEO hosting as a working method. Much like the people who suggest crappy hosts (Fatcow, iPage, etc) due to high affiliate pay, the SEO hosting companies will lie so as not to lose their gravy train. The companies selling SEO hosting have their interests as the basis of their comments, not yours.

Yes, search engines know which IP addresses are used by "SEO hosts". Another obvious parameter that betrays link schemes is domain registry data. Sometimes it's as easy as the technical content of the site (identical site code, for example). Again, there are quite a few vectors that can be cross-referenced to reveal a pyramid linking scheme. That's why this doesn't work.


Conclusion:

SEO should be a low priority for a website. Yes, it needs to be considered, but it's not the primary or even secondary aspect of a site. If you want a site to appear high in search rankings, it needs to be good. It needs to fill a purpose (niche or broad), and it needs a target audience (demographics). As search engines start to act more like AI librarians, and become harder to manipulate, more people will start to realize this.

If you put effort into developing the site, and not attempting to trick others, you'll do better in the long-run ... assuming your site idea was ever any good. If your site has nothing of value to contribute, it will not be indexed. This is not much different than a library not buying a book, because it's deemed to be junk.

It should be mentioned that site speed is now an SEO consideration. Many of these SEO hosts are lousy at speed, either because they're a branch of an already-poor mega-host, or because they're a poorly managed budget shop. It would be in your best interests to simply pick a good host. For example, HostGator, Stablehost, JaguarPC, EuroVPS or one of the others on our Top 15 Web Hosts List.


__________

* More on real marketers vs the fake wanna-be "affiliate marketers" --

If you'd like to actually learn about marketing, read this: http://www.amazon.com/gp/product/007...SIN=0078112060

The Core is an excellent college textbook, and it's done quite well to keep up with current trends in technology. (Even if it does start to get stale, the concepts are not going to change. Honestly, marketing concepts in general have changed very little in the past 50 years.)

Yes, it's about $100 new, or $75 used, but again ... college texttook! That's a typical price for a college textbook.

Marketing is about more than spam blogs, affiliate programs, "content" and "articles" (spam writings), and SEO. Those who actually understand the fundamentals of marketing are the people who can make a career out of running content-based websites. Those who want to use Wikipedia and Google searches as their primary source for understanding marketing are doomed to failure long-term, and many never see any degree of success.


__________

** More about Google's advanced methods for detecting bad content --

You may have noticed that this site has quite a bit of digital video articles and how-to guides, with a special concentration on video restoration and archival media. That field has heavy overlap with forensic media recovery, used for both law enforcement and historical archiving. And our site has a good bit of respect from professionals in those communities. But in addition to video, audio and photo, I have a decent understanding of forensic methods in general. No, it's not because I watch CSI on TV, but because I've studied it for many years.

In many ways, Google's search algorithms mimic forensic methods. You've likely heard of handwriting analysis (graphology), but that's actually a superficial observation of a document. If the document in question was a poem or essay, for example, you would notice not just similarities in the physical form, but in the content itself. The same words, dialects and grammar would be found. This is the result of your education, your geography, and other social/environmental criteria. For example, being from the South, I'll gladly use "ya'll" in a sentence, while otherwise adhering to the writing rules of my journalism education. While many people would think "okay, I'll just change my style," it's not that easy. In fact, aside from (1) being a psychopath or sociopath, or (2) being formally trained, it's impossible to do. You are who you are, and that comes through when you communicate.

In a digital world, you lack the handwriting, but the content remains. Based on what I've read in high-end printed tech journals, Google has implemented similar detections into the way it finds not just duplicate content, but semi-duplicate content and content written by link-scheme authors. Given how many web spammers are ESL (English as second language), picking out bad content is a relatively easy task for forensic search vectors. Many more are simply poorly educated, and view junk websites as a way to get rich quickly. However, their words and grammar easily betray them.


_________

*** More about content --

The biggest issue is that the word "content" (or "articles") has been altered by web spammers (a.k.a. "affiliate marketers") to essentially mean any barfed-down writing. I'm often reminded of junior high school, when we first started having to write one-page essays as homework. As a journalist, I've long been fond of writing, so such assignments were never an issue for me. However, my classmates were not always quite so willing. I remember arriving at school around 7:45 a.m. (class started at 8:15 a.m.), and sitting in the outdoor quad. There were always students sitting around writing their essays feverishly; most of them would get a C, D or F later that day during class. Quite often, the essays were factually wrong (maybe slightly inaccurate at best), or simply crap writing that managed to fill a page without actually making a statement. That's what web spammers do.

For probably 3-4 years, the "affiliate marketers" (who generally don't know tiddly about actual marketing) were getting away with high SERPs for their garbage. It was so popular that "content mills" (corporate-owned crap-writing factories) were founded, such as Demand Studios and Examiner.com. The most laughable aspect of these content mills was that the writers were often passed off as being experts when many of them couldn't even grasp basic grammar, much less the topic at hand. Existing sites like About.com and eHow.com became spammy and adopted the content mill strategy in order to increase their revenue and page views. (There has been a huge backlash in the journalism community, due to their attempts at hiring professional writers for unlivable low-dollar pay, but I'll leave that mini-novel for another day. CJR ran a great article on it within the past year.)

At the beginning of 2011, we saw massive corrections by search engines which have weakened their rankings, thanks to search vectors/parameters than can detect bad writing, as well as "keyword targeting".

Spamming the search engines simply does not work anymore, whether it's via differentiated A/B/C class IP addresses, or if it's by means of dumping random content into a site (be it auto-blogged, written for cheap by non-writers, or stolen from another site).


_________

More questions?

Feel free to ask more questions -- in the appropriate forum, and usually in a new thread -- if there's something you'd like to know. It's also worth mentioning that we go over some more advanced concepts in the Premium Member forum, both with past posts, or via new posts our members would like answers to. So a currently-$20 premium membership may be something for you to consider. (Premium Membership also removes some of the forum ads.)

I hope this has helped you to avoid unnecessary/worthless SEO hosting.

Skip the SEO host -- just get a good host: http://www.digitalfaq.com/forum/web-...-best-web.html

- Did my advice help you? Then become a Premium Member and support this site.
- Please Like Us on Facebook | Follow Us on Twitter

- Need a good web host? Ask me for help! Get the shared, VPS, semi-dedicated, cloud, or reseller you need.
Reply With Quote
The following users thank kpmedia for this useful post: Winsordawson (12-18-2011)
Someday, 12:01 PM
admin's Avatar
Ads / Sponsors
 
Join Date: ∞
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
  #2  
03-16-2012, 04:06 PM
kpmedia's Avatar
kpmedia kpmedia is offline
Site Staff | Web Hosting, Photo
 
Join Date: Feb 2004
Posts: 4,311
Thanked 374 Times in 341 Posts
Just to add something important to the previous post...

A quick tell that somebody doesn't understand IP space is that they use the term "C class" to describe the third octet in an IP address. Again, the abbreviation "IP" stands for "internet protocol" and nothing more. It's not something that can be, or needs to be, manipulated. IP address space has been classless -- technically referred to as CIDR, or classless inter-domain routing -- since 1993. That's almost 20 years ago; it's now 2012. And online technology years are faster than dog years. Something from 20 years ago may as well be from 200 years ago.

Beyond that, the C class was a specific range of IP addresses -- it DID NOT represent the third octet even in a classful network. To call the third octet a "C" class is about as accurate as calling it the "3" class or "gamma" class. You can't just randomly name it after the third identifier in some other random sequence. It's the third octet, period. The end. It's not a "C" class. Those who do so don't understand IP addresses. Even a cursory glance at Wikipedia would have made this obvious.

And therein lies the problem: Those who insist IP addresses influence search rankings -- especially due to manipulation of the non-existent "C" class -- don't understand what they're talking about. If they cannot even properly describe the basis for their theories, how accurate do you think their theories really are? (Not to mention the fact that major search engines have explicitly stated it DOES NOT work.)

If you want to improve your search rankings, by all means do so. Optimize the quality of content on your site, which will bring in both organic traffic (people) and inorganic traffic (search engine bots) -- aka "SEO". Trying to finagle the protocols that power online communication is not SEO.

- Did my advice help you? Then become a Premium Member and support this site.
- Please Like Us on Facebook | Follow Us on Twitter

- Need a good web host? Ask me for help! Get the shared, VPS, semi-dedicated, cloud, or reseller you need.
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
HostPolar Review ~ Linux VPS hosting, shared cPanel hosting kpmedia Web Hosting 2 07-08-2011 12:40 PM
Sony both 1st class AND 2nd class media? admin Blank Media 2 09-21-2009 04:06 PM

Thread Tools



 
All times are GMT -5. The time now is 02:51 AM