Do on-page factors really even matter? Check out my test results,

[Webmeister]

Recently I did a test using one of my websites that ranks highly for nursing- and healthcare-related keywords. I added links to about 300 websites using anchor text that was related to winning the lottery. Within six months, the nursing websites were ranking on the first page of Google for several of the lottery-related keywords. However, I have absolutely zero content on the website that is related to the lottery.

Please attempt to explain to me why on-page factors even matter with Google. Sure, relevant content on your website's internal pages is a boost - but could it be just as effective if the pages were on another website? I see a lot of people who get paranoid about onsite factors as if they were the make-or-break factors to search engine ranking. My tests have shown the opposite.

While I have your brain considering this issue, give me your thoughts on this trick question:

Which one of the following is better for SEO?

1. 100 relevant PR3 web pages on your own website linking to your home page.

2. 100 relevant PR3 web pages on someone else's website linking to your home page.

Thoughts?

[Webmeister]

Maybe a better question to ask would be this one:

Is there anything you can do to a web page's meta tags or content to trash its ranking on Google, or is its ranking determined 100% by off-page factors?

[mirrornl]

the 300 link to lottery where on-page
so it is not an off-page factor is it?

[tedster]

Is there anything you can do to a web page's meta tags or content to trash its ranking

I can guarantee you from the past week's experience that there is. I have a client whose highest traffic page fell from #1 on two big query terms to pages 3 and 4. What they did was change (not tweak, demolish) ALL the body content and the page template. I was not informed ahead of time and I only learned about it after the traffic drop appeared.

Their site uses 6 templates, and they switched the same URL from the existing template to one of the others that they use for new announcements and promotions. They even kept the same title and H1 tags, but they did switch the big Flash movie that heads the page.

Traffic loss? 28,000 to 100. The take-away is this: on-page factors still influence rankings. This internal page still has backlinks at the six figures level with a natural percentage of keyword anchor text.

[buckworks]

28,000 to 100? Yikes, that would be painful!

Ranking depends on a myriad of factors, and the way they combine probably matters as much as what any particular factor happens to be. It's hazardous to point at any one thing and say "this matters most" or "that matters least".

Everything matters, including many factors we can only guess at.

[TheMadScientist]

the 300 link to lottery where on-page
so it is not an off-page factor is it?

I think you're splitting hairs about the wording rather than going with the point personally... My guess is this would not be reported if the links were not on the 300 websites going to the page on the nursing website, even if it could be read differently. IDK why people feel the need to do this here as often as it happens rather than just saying thanks or going with the point, especially when the OP is giving away some good information and insight, but each to their own I guess.

Personally, I think it's a great question and I agree with tedster on the meta data and the point there is a much better chance for on meta data to be a minus than a plus in most cases.

I see your point on the links, but have sites where the opposite is true and they don't have many links (any except internal and some with 'non-keyword text links' to some ranking pages) but still have pages ranking highly, although for fairly uncompetitive terms, so I would say on-page factors play a role, but definitely think the linking situation is interesting. Thanks for sharing.

[Webmeister]

I can guarantee you from the past week's experience that there is. I have a client whose highest traffic page fell from #1 on two big query terms to pages 3 and 4. What they did was change (not tweak, demolish) ALL the body content and the page template. I was not informed ahead of time and I only learned about it after the traffic drop appeared.

Their site uses 6 templates, and they switched the same URL from the existing template to one of the others that they use for new announcements and promotions. They even kept the same title and H1 tags, but they did switch the big Flash movie that heads the page.

Traffic loss? 28,000 to 100. The take-away is this: on-page factors still influence rankings. This internal page still has backlinks at the six figures level with a natural percentage of keyword anchor text.

Did they change the entire site? If so, then they probably changed pages that have links pointing to your home page on them. If that is the case, I believe it was because the internal pages linking to their home page were modified - not because the home page was modified.

If CNN.com changed its template completely, would that kill their #1 ranking for "headline news"? No, because 99% of their backlinks are on other websites, and it wouldn't matter. I'm guessing your client probably trashed pages that affected the majority of their backlinks.

[tedster]

They just changed the content area for that one URL. No site-wide changes at all.

Almost all of their 700,000 pages are exactly the same as they were before the marketing department got this particular brainstorm and ran with it.

This is not exactly the same as your test - showing that off-page factors can generate rankings for content that is not on-theme. But it definitely shows that on-page factors still do matter, even for a major enterprise with backlinks to burn.

[tedster]

Your closing question is a good one, by the way. I'd say that 100 external links will outperform 100 internal links, if the externals are from different domains. If the externals are just ROS links, then I'd go with the 100 page internal linking.

[1script]

@tedster:

I can guarantee you from the past week's experience that there is. I have a client whose highest traffic page fell from #1 on two big query terms to pages 3 and 4. What they did was change (not tweak, demolish) ALL the body content and the page template.

This is very relevant for me, so I'd be so happy if you could provide a few more details :)

First off, I am the first to admit that I am prone to doing what your client did and sometimes even couple times a year ... Every time a nicer WP theme comes along, I am tempted to download and install it on my blog right away. The thinking was (until I read your post that is) that the pages are going to have the same backlinks, URL and structural elements (title, h1,h2 etc.) as well as content, so why should it matter so much?

So, are you at liberty to divulge the difference between the old template and new in your client's case? Say, H1 moved 20 lines down or up, a block of JS moved up or down, content moved up or down, stuff like that. Any additional detail will be greatly appreciated (and probably save me from pulling too much of my hair out the future :)

[Webmeister]

Thanks for the input Tedster - it helps me brainstorm this issue.

Let me give you another example of a recent test I did. I purchase a domain that had a PR3. It ranked on Google at #9 for a popular phrase that brings in some good AdSense traffic. I changed the entire website - template, links, page content, etc. and it didn't affect its Google ranking at all. However, I made another change that dropped it completely out of Google: used 301 redirects to point the non-www version of the site to the www version. Within 10 days, Google dropped it completely. Three weeks later, I took the 301s down and allowed the non-www version to be active - and within a week, it popped back up on Google at #7 for its keyword phrase.

[TheMadScientist]

The last one is really interesting, did you try going the other way too?

I mean a www to non-www with a redirect? I'm wondering if G had determined a version of the site to rank and 'separated' the other version of the site making it seem 'new' or something when you redirected to it? Either way, that's really interesting.

[tedster]

They have two templates that we can call "landing page" templates - they're designed for top-level entry traffic. Both have an interactive Flash movie on the top left and both have a solitary, H1 element to the right in a very big font. Main navigation is under that top area in a narrow strip. The navigation remained unchanged, the title and H1 both remain unchanged.

Then the content area below that large header and nav is where the change happened. It can be either four content blocks (2x2) that each end with a link into the deeper content, or just one marketing blurb paragraph. They changed both the Flash movie and all it interactions, and they switched from the 4-block informational template to the single block marketing-blurb content, with completely new copy and just one internal link compared to 4.

28,000 to 100? Yikes, that would be painful!

Yes indeed -- but with a certain type of major enterprise, they are just now figuring out the web. So they don't really feel the pain the same way another business might.

Since I pointed out the issue, they've asked me to right up a case study for internal distribution so that they can correct their internal processes. Now THAT is a savvy enterprise. If they keep that attitude, they will slowly but surely do a lot better with their web property.

[1script]

@tedster:

So, in summary, it's not really the template they changed but the content itself. I guess, it's reasonable enough to draw a line between on-page factors such as layout, h1, title, metas and such and the content itself. I'm not sure if we call content an on-page factor. Do we? In my mind content is THE page itself. Or, rather a combination of content and its URL is what absolutely defines the page. Everything else appears to be secondary to those two.

So, I guess, it's not really that strange that they dropped so hard - they completely changed the content.

Bottom line - if you have something new to say, create a new page. Case study written? ;)

[Sgt_Kickaxe]

#2 is better, Google has this thing about trusting what we put on our own websites.

[dazzlindonna]

I kinda have to agree with mirrornl and I don't think it's splitting hairs or being unreasonable to point it out. The links on those pages are actually content on those pages. I'd say that the nursing site he's talking about must have enough trust and power that small bits of content placed on the page can rank - and yes, those links are small bits of content. OP did say that "several" of the lottery-related keywords were ranking, so to me, that means "not all of them". If the "several" were not overly-competitive, and the nursing site has enough trust, I can see how adding "links which are also content on the page" could cause the site to rank for those phrases.

[TheMadScientist]

The links on those pages are actually content on those pages.

That means they're off-page factors for the page that's ranking.
They're not on-the-page ranking, are they?

The OP's talking about links on different pages (they're off-site factors too, because they're not on the same site) having more effect than the content on the actual page (they're off-page factors in the world of SEO) and the fact the links have nothing to do with the topic of the content on the actual page ranking.

The content and links and 'stuff' on the page is considered on-page.

Anything not on the page ranking (which definitely includes links on other pages going to the page) is considered off-page.

It's really not that difficult to understand and there's really nothing to argue about if you know the terminology. If you want to try and change it to add some extra confusion to things or 'develop your own language' and confuse people who don't know better then fine, but on-page is what's on the page and off-page is what's not on the actual page if you talk with people who know anything about SEO.

A really easy way to understand the difference between on-page and off-page is to simply view the source code of any web page and print it out. If it's there it's an on-page factor. If it's not there and influences the rankings of the printed page it's an off-page factor.

[tedster]

Not sure what the difference of opinion is, here. But I do know that many webmasters downplay the idea that a link and its anchor text is an on-page factor for the URL where it appears - as well as an off-page factor for its target URL.

The focus tends to be on the link's target page, at least until the host page starts to rank for the anchor text involved. This is one area where Google's dials keep getting turned a lot. The balance of sending traffic to the higher level (but linking) page or the deeper target page that is dead-on for the sub-topic.

In the case of my client, they removed three of the four content area links and their anchor text. This, in my opinion, was a stronger factor than revising the regular text copy. However, in this moment of G-Time, it seems that major revisions of any kind can give the page or site a "time out" for getting to boisterous or something.

[TheMadScientist]

What you're saying is exactly what I'm saying, so I don't know where the difference of opinion is either. Like you say, the link and anchor text is on-page where it appears and off-page for the page it links to. Seems simple and I didn't know there could be (is?) confusion about it, or a need to discuss it, but since it was brought up as an 'error' from the wording of the OP, I think for future readers it's important to try to eliminate confusion... IMO this is a 'tough enough' arena for competing without having to wonder if on-page really means on the page it's on or if it really means something else, so I figured I would try to clarify what is AFAIK generally accepted terminology in this line of work.

[Lapizuli]

Oops, I'm lost regarding the first post. Webmeister, are you saying:

1) On your healthcare website, you added 300 links on numerous pages (not just one) to numerous other (healthcare related, not lottery related) websites that are not your own, with variable text suggestive of the lottery? Like "Win big at the sciatica lottery by trying this therapy"? And then the nursing website took off in lottery-related keyword rankings based on lottery anchor text alone, even though neither the host pages nor the destination sites were about the lottery? And (you didn't say) presumably it didn't hurt your healthcare-related keyword rankings?

Or am I interpreting that incorrectly?

And

2) Is this sentence: "However, I have absolutely zero content on the website that is related to the lottery" referring to the nursing website having no lottery-related content, or to a website you have about the lottery that has no content on it yet?

Sorry, my brain's just foggy this morning. I have a feeling I'd find this conversation awfully interesting if I could figure out ground zero...

[mirrornl]

We are not making our own fancy language
sorry if i misunderstood (English is not my native language)
and it's not a biggy...
but OP says:

Recently I did a test using one of my websites that ranks highly for nursing- and healthcare-related keywords. I added links to about 300 websites using anchor text that was related to winning the lottery. Within six months, the nursing websites were ranking on the first page of Google for several of the lottery-related keywords. However, I have absolutely zero content on the website that is related to the lottery.

so i understand he did put 300 links on the nursing-website, and that same website is ranking now for lottery searches ?
wich is an on-page factor?

excuse me if i'm wrong

[TheMadScientist]

No, that's where I was saying to go with the point of the post.

They put (to my understanding) 300 links related to the lottery on different websites pointing to a page on a health-care related website without any information related to the lottery and the page on the health-care website without lottery terms or information present about the lottery started to rank for the lottery terms.

Since this happened, the question: Do On-Page Factors Even Really Matter? was raised, because the page ranked completely from off-page (external) factors for terms unrelated to what the page is actually about.

[mirrornl]

OP must clarify

I added links to about 300 websites

i understand he did put link3 to 300 sites,(pointing to)
not on 300 sites (not added to 300 sites)

[TheMadScientist]

That's where subtle differences in English and possible interpretations can be really difficult to 'get'.

You are thinking the OP is saying they added the links to (in the direction of) the 300 websites, but it can also be read as added to (on, within) the 300 websites... It can be read either way, so IMO, go with the point.

I'll let the OP respond if they feel like, because it can be read both ways, but the context clues of the post and paragraph indicate the links were added to (on) the websites, not in the direction of the websites. They were put there (on the 300 websites) in addition to the links already present... They were added to those sites, not 'in the direction of' those sites based on the context of the sentence and the post.

English is a highly contextual language, so don't feel bad or anything, but remember there are often two ways to read something, so to really understand what someone is saying (or trying to say) you have to go with the context as well as the actual words used in many cases.

[edited by: TheMadScientist at 6:17 pm (utc) on Jun 27, 2010]

[Lapizuli]

Thanks, TheMadScientist & mirrornl.

The other part of my confusion comes from:
"test using one of my websites that ranks highly for nursing- and healthcare-related keywords."
and
"Within six months, the nursing websites were ranking"

My brain's in linear thinking mode this morning and can't wrap around the logic until I've resolved the rest, aargh!

[TheMadScientist]

Of course, it could be I read it wrong too, but in 'going with the point', as many times as I read the post it seems to say some links were added somewhere (obviously pointing to a site(s) from another site(s)) and the page(s) the links were pointing in the direction of were unrelated to the link text but began to rank for the terms in the text... That's what I get out of the post, and even though I can read it 3 or 4 different ways I don't see where the OP says anything about the page(s) the link(s) were placed on ranking for the terms... I get the opposite every time and however I try to interpret the exact wording.

The context (even if not completely understanding the exact words) says to me links were added somewhere, and the unrelated page(s) they pointed to started to rank for the terms in the links not the text actually on the page the links were pointing to.

[mirrornl]

I can see it also the way you are reading it TheMadScientist
i was interpreting it my way, because i had the value of anchortext in outgoing links onpage in mind

[1script]

Isn't Webmeister describing a classic case of a Google bomb? Given that Google is so link-dependent in their algo, how could it not work?!

I have to give Webmeister kudos for going through with such a dangerous experiment (except of course if the nursing site was not in fact his :) ) - 300 lottery related links, especially if added at the same could have brought the site down.

I'd be also interested to learn what type of link were they - ROS, homepage, in-content, do-follow, no-follow, comments, forum posts etc. I'm sometimes finding it hard to get 3 good links when needed and some people are wielding 300 just for an experiment ...

[dazzlindonna]

I've been around the SEO block a few million times, and I certainly understand what on-page and off-page means. We all agree that the page on which the link and anchor text appears is on-page and the the target of that link is off-page. What I and mirrornl are saying is that if the pages on which the links are placed (on-page) are now ranking for the phrases within those links, then on-page is obviously at play here. I don't believe the OP has given any of us enough information to determine which pages are now ranking, but it's certainly possible that on-page is playing a factor if the host pages are the ones ranking. So, no, I don't think anyone is reinventing the terminology, or even misinterpreting nuances of English (I'm a born and raised American who majored in English). I do think the OP was vague enough to have allowed different possible interpretations because we don't know which pages are ranking for which phrases, but I don't think we need to throw out the possibility of on-page being a factor until we know that for sure - are the host pages the ones that are ranking? Or not? You read it as the context being that unrelated pages they "pointed to" started to rank. I didn't read it that way. Sure, that's the normal way of thinking about it, but that's not what the OP said. The OP wasn't clear so until that's cleared up, I don't think anyone should be scolded for interpreting it differently. If the OP comes back and says it's the target pages that are ranking, then that's an off-page issue, but if he comes back and says the host pages are ranking, then it becomes an on-page issue.

[buckworks]

vague enough to have allowed different possible interpretations

Avoiding that kind of thing is why good technical writers can command such high salaries! :)