Google "distinguishes natural links from unnatural links"?

In order to "find" something, you must first define it. That's the part that worries me. How is "unnatural" defined?

None of the suggestions I've yet seen for doing this make any sense unless "natural" is a synonym for "relevant" - and if that's what meant, why not just say "relevant"?

 

With these kinds of claims I think google is inviting its early death.

 

A question I am pondering myself and I am also thinking will they tend to sway towads a directory as being unnatural? I know I would.

 

The best I can tell the real definition for "natural" vs. "unnatural" links in the Google database follow a system of percentages.

Let me illustrate this idea by example:

• If a site has 1000 links pointing to it, and all 1000 links have the same imbedded text in the link, then there is something "unnatural" about that. (It looks like all of the links were created by the marketing team of the domain in question.)
  
• If a site has 100 links pointing to it, and those links take the form of ten different imbedded keyword text in the link, then that is more "natural". (It appears on the surface that individuals are doing their individual part to recommend websites to others.)

It is possible for two or ten webmasters to use the same descriptor (imbedded keywords) to describe a website. But this will not occur 100% of the time.

When one descriptor is used in 80%-100% of the links to describe one domain, that is "unnatural" and will be penalized by Google.

The difference between "natural" and "unnatural" is simply a matter of percentages. Google is playing the odds to figure out what is "natural" (trustworthy) and "unnatural" (untrustworthy).

If you want to beat the system, keep changing your imbedded keywords in the links that you create pointing to your own website.



p.s. I also agree that Google is definitely not penalizing links from the same IP block or address. Blogspot.com accounts all reside on the same IP address, and yet, blogger accounts get good results in Google regularly. Those subdomains receive and give PageRank to others.

The key is not the IP block, but the subdomain.domainname sequence.

Google does treat subdomains as separate entities at this time.


Bill Platt

 

Thanks Bill - I agree with your description of natural vs. unnatural (from a theory perspective - we have no idea what Google is doing, but this makes sense).

I am not so sure about the IP block, as Google owns and operates Blogspot - so they can control how those sites are perceived in IP context.

 

I think its basically Google wants to do what they want to do. One thing is for sure though that Google doesn't want people to know what they are doing. So maybe they are acting and making people confused with whatever they claim.

 

I do realize that the Blogger/Google relationship might make a difference on how Google treats that IP block. But, I have also seen additional evidence that indicates to me that Google is not yet counting same IP addresses as a negative factor.

This may change, and I fully expect it will at some point, but I honestly don't think it is a real factor just yet.

 

I don't agree - I think that's yet another "Google myth".

If 1000 people all link to Apple Computer with the anchor text "Apple Computer" or Microsoft with the anchor text "Microsoft" or Google with the anchor text "Google", do you really believe those links will be discounted as "unnatural"?

 

I think Goolge worrying about natural and unnatural links is not near the top of their list of important things to do. Spam is a huge problem and MFA sites, my guess is that they are concentrating more on those. I think they will still discount links from link farms or FFA's but other than that I doubt they care too much.

 

This is a toughie, because we're just guessing what Google means when it says "natural". I suppose they don't mean links created by mutant zombie aliens, but what's the alternative? Links created by webmasters. Human webmasters, who have biases, opinions, favourites, and a need to make money. No link is divorced from those things.

My take on "natural" linking is that it matters how many people placed a link, and Google are looking for patterns that suggest one webmaster or entity is behind a set of links, rather than lots of people.

 

I agree. I don't think google can really determine natural from unnatural based on anchor text. Another one would be 'ebay'. Most sites that link to ebay use the anchor text ebay.

The majority of natural links my site receives (sites that link to me on their own) use my site name as the anchor text. So, are we saying if my site name is used as anchor text too much I will get penalized for this or the links won't count. I highly doubt it. Although, who really knows.

 

Let's just say that if i created a search engine algorithm i would include finding a way to distringuish between natural and unnatural links.

What search engines actually do, i don't know.