Name Servers and SEO
 

If a host wants you to use a custom name servers like ns1.yourdomain.com and then for all 30+ of your other websites to use the same DNS. They do this to make it easier for them which I understand but I personally would rather use the web host DNS.

Where do you stand on this as far as SEO benefit?

name servers are not so much an issue as the IP addresses on which the sites reside as far as I know.

That would be incorrect.
:2 cents:

enlighten us then

unique DNS per IP.
Just saying.

I've spent alot of time studying search engine algorithms. I have yet to see an authoritative example of name servers affecting ranking in a significant way. I have however seen IP addresses of websites and even geographic location of IP addresses/hosts affect ranking.

If anyone has any links to any authoritative references to name servers affecting ranking, other than issues like the availability of name servers or the performance of them affecting ranking, then I would like to see them.

What's the point of that? I've never seen a nameserver have ANY effect on SEO.

You've probably never saw the point of using different registrars, or Analytic accounts.
:2 cents:

i agree .....

If you have links to actual examples or papers which demonstrate name servers adversely affect search engine rankings then please post them, otherwise it's just talk.

The issues related to name servers that can affect ranking, as per my understanding are:

1. Uptime / availability of the name server
2. Correct configuration of the name server.
3. Distribution of the name server - goes to configuration - eg: if you have two name servers ideally they should be located on two separate networks.

If the name server is not always reliable, or slow, this results in the crawler seeing your site as slow to load, affecting ranking. If the name server is incorrectly configured then it may affect ranking as the crawler may see old cached entries or incorrect serials. If your name servers are clumped together then any network issue between them and the crawler will also affect load time or site availability which can affect ranking.

However if your name servers are dispersed and working properly, then what they are called should not affect ranking. Also having one set of name server instances should not affect SEO.

I'm happy to be shown where I am wrong here.

I've found that when using the DNS of a webhost with "serious" past infractions, you're main keyword's can take much longer to gain traction....but no other real "penalty" in the long run as your site gains the trust factor. This is especially with some of the higher-end/more sought-after keywords. However this is just my personal experience from working with several 50+ domain networks.

Then again, this can also just be IP based & have nothing to do with the NS. So even when using your own NS's, you're still using an IP for your NS that's on their network.

Of course there are some major benefits to using large-ISP name servers when creating multiple domain sp4m n3tworks. Which is exactly why it's taking longer for websites to gain traction using the larger/cheap hosts these days.

I see no real difference when using my own name-servers for my main site's (domains that are a niche of their own, user-generated content, no-cookie cutter stuff). I also notice a markedly faster resolution with my own properly configured DNS server/s for my network/s, especially during peak hours. By this I mean multiple name servers, setup in multiple locations, on multiple servers, for redundancy. (I'm sure you know all about this....this is simply for the benefit of "new guys") It's basically a personal version of UltraDNS that I've setup for some associates.

So if you want a real benefit from using a custom name-server. Make sure the IP's actually reside on a fairly trusted host...not the latest "hit & run" spot. Also make sure to ask if they have recently purchased any new IP space in the past 2-3 months...always a good practice. :2 cents:

Yes pr0 is absolutely correct here, the history of an IP address can be an issue, especially if that IP has been associated with link farms or spam.

From their patent:

40. The "expiration date of the domain", the "domain name server record" and the "name server associated with the domain" are all parts of how Google will establish the legitimacy of an "associated" site.


So Google states they do use it. However, they may have stopped or gave it .0000001 on the scale of who gives a shit for now.

And if you're going to configure your own DNS/Nameserver network. Make sure the servers are diversified correctly & in/or near major internet backbones.

I.E.
Chicago, Los Angeles, Miami (example: Mojohost), DC region + Ashburn/Reston VA, New York City (example: Webair), Atlanta (example NationalNet) & Dallas, TX.

And if you're making real money from euro/asian/south pacific traffic.

Euro-zone UK, Netherlands 'Amsterdam of course' ( example: webair), Sweden (excellent infrastructure); Latvia, Czech Republic & even Romania.

Asia: South Korea, Japan & Hong Kong (They all have outstanding infrastructure to cover most of the Asia-pacific region) So choose the best 1-3 to suit your traffic needs.

Australia: you've got New South Wales, Victoria & Melbourne if that's a main market for you. However I don't know too many people expanding their name-server distribution into Australia yet. I would imagine something like UltraDNS might have an AU a presence.

South America: Although a typically Miami is used by most networks, you may want a presence in an actual South American country with real infrastructure in order to drop an extra few milliseconds. I would only suggest Brazil, Panama, Argentina....

Also if you're not happy with your LA presence (for S.American purposes), you can get a decent connection Mexico these days, which would make you one step closer to S. America without having an actual presence in the deep southern regions. However I've always experienced extreme latency issue's in MX. Mostly time-of-day "rush" dependent but not always.

Hope this shed some light for a few people :pimp
I'm out!

We use a different NS for every Class C subnet. We have clients that do it for each domain/IP. Definitely beneficial if you are trying to be discreet.

:winkwink:

Some quality answers and thank you!

AdultKing: My advice: forget about evidence that A or B or C effect your ranking.

Google (and other search engines) are in the business of indexing information. They take tons of data and filter out small amounts of information that they think is relevant to the end user who performs a search query.

As outsiders we have no way of knowing how exactly their algorithms work. We can make educated guesses, but we can never know for sure. We know they look for certain patterns (like for example the amount of inlinks), but we can never know exactly what patterns they are looking for, what kind of information they currently take into account and what information they might take into account in the future.

SEO is not about linkbuilding and writing content per se, it is all about managing information. Controlling as much as possible the information that is available about your sites. Hiding the fact that one person or organization owns and controls 100s or 1000s of sites in a certain niche. Manipulating the info out there (linkbuilding = manipulating data)...

If 500 sites each use a different IP address but they all have the same whois info or regdate or use the same nameservers or are registered with the same registrar or all have the same adsense cod eon them or all run the same cms or the same webserver with the same modules... than that is information Google might use to determine that those 500 sites might belong to the same person.