From my experimenting on and off page factors definately both still have influence.

I recently did the following:

1) improve in site links using keywords
2) generate more incoming links
3) changes to the pages in the site

Each of these three actions had definate effects, improving our rankings.

<OTWarning>

I don't think anyone should be scolded for interpreting it differently.

Well, maybe if what you had to say was said in a amicable, friendly, clarifying way, rather than a condescending, know-it-all, 'I know what you mean by your post better than you' way the replies would less 'scolding'?

For instance, what if you said:
Could you clarify if the links were on the pages ranking or pointing to the pages ranking?

Rather than:

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

If you had I probably would not have tried to stick up for the OP a bit, or am I maybe I'm misunderstanding the condescending tone of your post? If so, and your post is not a condescending quip at the person who actually knows the situation, please feel free to clarify what exactly it is, because it's certainly not a friendly question for clarification in any way shape or form, and there aren't too many different ways I can see to read yours.

Should I split hairs on whether you meant where or were in the first line of your post, because I probably could?
</OTWarning>

Back on topic:
Maybe we can all agree whatever page these links were on, whether the ranking page or the page pointing to the page ranking there was NOT MUCH (a few words at most) text in any way related to the terms the page was ranking for and it should not, based on the overall content of the pages ranking, have ranked for those terms because all but a few words (probably 99% or more) of the content was about an entirely different topic, which seems to say 'on page content' in this situation seems to be heavily discounted?

AFAIK, the text in links to a page still 'count as part of the content' of the page they are pointing to, so whether the links were on the pages ranking, or pointing to the pages ranking, AFAIK, from a 'purely Google perspective' the pages ranking should have been 'credited' with the same amount of 'text on the page' either way the links were pointing, and possibly more if text surrounding the links is also credited to the page the links are pointing to.

The point still seems to be the same, very little to no content on the subject, except some link text credit one direction or the other and the pages ranked for the terms. I still keep getting an 'essentially the same' view of the situation, even when I try to interpret it totally differently and look at it logically... I keep coming up with little to no text on the subject the pages were ranking for 'credited' in any way shape or form to the pages ranking, so do on-page factors matter any more seems to be a valid question, to me.

If you look at the two possible situations and ask a simple question or two IMO everything makes more sense and it really doesn't matter which page the stinking links were on:

### # ###

If these links were on the pages ranking, why on earth should those pages have outranked pages with content on the actual topic if on-page factors were counted in the rankings?

You've got a page about nursing with some links on it containing text about the lottery; You've got a page totally dedicated to lottery information.
Based on on-page factors, which should rank for the lottery terms?

### # ###

If these links were not on the pages ranking, but rather pointing to them, why on earth should those pages have outranked pages with content on the actual topic if on-page factors were counted in the rankings?

You've got a page with some links pointing to it with text about the lottery in the links. The actual topic of the page is nursing; You've got a page totally dedicated to the lottery. Based on on-page factors, which should rank for the lottery terms?

### # ###

Either way, the on-page factors of any page completely surrounding the topic should have heavily outweighed the on-page factors of the pages noted in this thread. It doesn't matter which direction the links point, because the on-page factors for the topic on the pages in question have to be much less than any page completely dedicated to the topic, don't they?

So do on-page factors really even matter any more?
Seems like a really good question to me.

And I think you are a Mad Scientist :-)

I just don't follow your post, it is perhaps the way it is written, sorry but I can't make head nor tail of it.

And I think you are a Mad Scientist :-)

LOL, me too...

Basically, I'm saying the on-page factors of any page totally about the lottery should far outweigh any on-page lottery term factors on a page about nursing.

A page totally about the lottery will logically, reasonably have:
Title about the lottery with some type of lottery terms in it.
Links about the lottery on the page to other pages about the lottery and lottery information, and text in those links about the lottery or using lottery terms.
Text on the page about the lottery with lottery terms in it.

A page about nursing/health care, either with inbound links containing lottery terms, or with outbound links containing lottery terms only has those terms for any type of page factor (on or off) related to lottery terms... It will logically, reasonably have a Title about nursing or health care. Links to other pages about nursing or health care. Text on the page about nursing or health care.

If the on-page factors are really weighted in the rankings, then how could a page with essentially no on-page factors possibly out rank a page with 'all the right on-page factors' for the same terms?

It's like you're saying about my post:
It makes no sense...

Here's another question along the same lines:
How can on-page factors be deemed 'important' or 'critical' when they can be overcome by a page with essentially no on-page factors for the terms it is ranking for?

Again, it makes no sense.
If the on-page factors were truly 'critical' then they should not easily be overcome by a page with essentially no on-page factors related to the terms, right?

This is my first post so I apologize if I'm stepping on any of the guidelines.

I would like to say my scenario was not a test but was a cause of some additional content being added and the most important part I think some "internal contextual hyper links" being removed. We had not planned on removing the hyper links permanently but my head editor had planned on circling back once all the content was added and replacing the "internal contextual links" after all content was added.

We started adding new 100 % original unique content and over a 3-5 day period, I was monitoring one of our top 3 keywords, which is in a fairly competitive space (About 177,000,000 results (0.27 seconds)). Upon checking around day 5, we had a 2 page drop, this drop was from page #1 spot 7-8 to the last spot on page 3. This was puzzling to me, so I tried to see if google had cached the latest version of the page or not, I noticed none of the pages had been cached with the new content and missing hyper links. Well as we all know the cache date isn't always the day in which you are looking, meaning that when I checked 2-3 days later the pages had a cache date on or before the original day I was checking (hence google had seen the changes). I'm right in the middle of waiting for all of these pages to get re-indexed with the new content and the replaced contextual links re-added so I can't tell you if adding the links back have fixed my "DROP".

What I can say is the only things that were changed were the following
-Added new content to each of the pages approximately 250-300 across 350 ish tier 3 pages
-The content was more relevant as me moved from boiler plate content to 100% relevant content for each page.
-2 contextual links were removed from each of the 350 ish tier 3 pages. #1. one that linked back up to the immediate parent category and #2. one that linked back up to the grandparent category for lack of a better term. The grandparent page is the term that dropped.
- everything else monitored stayed exactly the same

just offering my recent changes that might add to the original post. If in the forum/thread I will be happy to update all of any changes.

I find this topic very interesting. In my experience off-site seo can cause a site to rank for terms that are not mentioned on-site as OP mentions. It is unsual since Google tried to address Googlebombing, but it still can happen. Off-site links that seem unnatural, thoug, can hurt your rankings (in my opinion and I know this is controversial - but I think Google misdirected the truth by saying they don't penalize sites, but instead they seem to use the term "filter"). In my experience, unnautral means multiple similar anchor text exactly matching on-site seo (title/H1). Also, run of site links can kill a site they point to for that exact phrase - or at least it has for me in the past but it was more than a year ago.

I think on-site seo can also hurt you if overdone. In fact, I wouldn't be surprised if a competitive term got ranked better on a site with no on-site mentions of the term but very good off-site seo. In the past I've had pages rank markedly better when I significantly limited keyword density on very competitive topics.

I would rather have good, natural off-site seo any day over increased internal links.

Having said that, I noticed for a while in early May off-site seo was signifcantly turned down as a factor but that seems to be "fixed" the last 2 weeks. Lots of adjusting lately but it does seem to be settling down.

It actually wouldn't surprise me too much, especially after reading through differing results in this thread, if 'on-page factor points' range basically from 'zero to minus' and 'off-page factor points' range from 'plus to minus', with 'on-page stability' being one 'higher weight' factor.

This is the best I can put what I have in mind in words right now, so please don't go with the exact wording, but more with the concept of on-page factors being a more probable 'nothing' to a 'negative' and off-page factors being a more probable 'positive' but able to be discounted all the way to being a 'minus'.

Yeah, text on the page = found the text over here!, so it obviously counts in some way, but beyond 'Hey, we found text' I'm not sure how much where or how it's formatted or what tags it's within really matters any more, because I'm seeing very little evidence of that posted beyond the 'it's text and it's here' factor, so in my mind that's not really a plus for rankings.

I guess what I'm wondering is if on-page basically 'gets you in the game' and could possibly take you out, but I'm seriously starting to wonder if it's really scoring pages higher. It might help you score in more places, but I'm really wondering right now if it is actually the cause of 'scoring higher' or if once you get the 'Found Text!' point that's where on-page is basically ending? *

* Just thinking while I'm typing right now, so don't take any of this as 'solid fact' or 'solidified idea'. Right now it's much more like 'thinking out loud' and subject to change at any time.

Actually, when you said to me, "Well, maybe if what you had to say was said in a amicable, friendly, clarifying way, rather than a condescending, know-it-all, 'I know what you mean by your post better than you' way the replies would less 'scolding'? " you should know that I wasn't the one who made the statements that you consider to be condescending. Someone else did, and as far as I can tell, that someone else doesn't have English as a first language. I doubt he was being know-it-all, but I could be wrong. I felt, in your scolding of him, that you were the one that was being know-it-all. Each to his own interpretation, I guess.

Sorry I don't go back and re-read all the threads I reply to and thought it was you...

I could have put 'if the person' rather than 'you' but IMO the point is still the same, isn't it?

If people are friendly and amicable, then cool, I'll do the same, but if the post I was responding to isn't 'trolling at it's finest' I guess I don't understand what trolling is... 175 posts in 4+ years. 2 lines. Telling the OP they don't know their own situation.

If that's not trolling, please, explain to me what exactly trolling is?

Honestly, I didn't know it wasn't you who made the original post and apologize for thinking it was.
If I had realized I would have worded mine differently, even though the point still remains the same.

Just as tid bit from my previous post, what i thought might be an over optimization error with internal contextual linking kinda of snow balled in to multiple issues. I think there is a tighter filter on contextual linking in regards to any page in your main navigation. I'm not saying it is the secret sauce, but here is what I have noticed across both of my business's and their sites. I had done some internal contextual linking, again some, not over the top. I moved to a cascading navigation for usability, trying to make everything one click away for my users (yes my users not Google's). I didn't think to remove the (before main nav change) contextual links or to stop doing what I had always done. My results are this on both sites, 2 different verticals one niche one high traffic my 1 main trophy phrase and a few tier 2 phrases (now in the cascading nav) dropped to page 4, and yes spot #1. This confused me as we continually were adding more content for our users. I originally thought it was the new content, but as I continued to dig and remove I noticed that each site had been hit by the same thing/filter/penalty, again they are at 2 totally opposite ends of the spectrum. I started to notice by checking the site:site-com keyword command and these pages were gone, and my home page was showing instead. I started to see that the pages that had dropped where all in my main nav, in both instances. I have removed as many internal contextual hyper links as I could find, and the sites are now both back on page 2 and page 3 respectively. I can't tell you the threshold, but as I started to look at how this could be useful to my user I started to realize that I had done it simply for rankings. I mean it was in my main nav and footer or one and my main nav footer and side nav on the other. So yes I got greedy I guess. Not sure if it helps, but I do think Google tightened the screws on that filter.

In both cases anchor text, nav text or nav alt were all identical.

In both cases anchor text, nav text or nav alt were all identical.

Interesting observation but, in trying to wrap my head around it, I admit I need a few clarifications regarding your contextual links implementation:

Are you saying that significant amount of your internal pages were linked to from other internal pages, not just one-to-one but many-to-one? And for each "landing" page just one anchor text was used on multiple linking pages?

Given that you call the links "contextual" I have to assume the anchor text of the links was an integral part of the text of the linking pages. Are they very similar in content then?

Also, in your quote above, what do you mean by "nav alt"? Are you saying some of the links were actually click-able images? How are they contextual then?

Anyways, this discussion is of interest to me because I'm also doing a lot of interlinking using "Similar Posts" feature and I'm also struggling with some of the best KWs pushed to page 4 of SERPs. I would not call my links "contextual" though because they are not quite inside the text of the page but rather in a clearly defined area marked "Related Posts". However, anchor text is similar to the text used on the page (hence "Related") and it is also always the same anchor leading to the particular linked page because its title is always the anchor. So, maybe there are some similarities here.

1script - Sorry if I wasn't as clear on the first go as I think I posted last night at 1 a.m..

The reason I said nav text or nav alt was beacuse this was on 2 completely unrelated sites and one has a main nav of images and alt tags and the other has text in the main nav.

In regards to the contextual links - these links were within a body of content (300-500 words). Within the content I hyper linked sometime ago the keyword "used widgets" to the used widgets page. (none of these were to my homepage FYI)

Now fast forward without realizing this could possibly be an issue I moved to a cascading main navigation for each of my main sections of my site. The new navigation dropped open giving users the ability to easily get to the sub-categories with in one click via the main nav. So lets just say I went from 5 main nav destinations to 25 (5 x 5). What I didn't realize until 30 days of research was that each of the pages that was added to the cascading nav, that was internally contextually hyper linked to with the exact same keyword "used widgets" had dropped the 3 pages. This only came to light because they are(were) easy top 5 most powerful pages on each site, as well as without a doubt the most powerful page on my site for that exact keyword. I didn't incur any site wide penalties, but specifically these pages did drop independently. That being said I'm sure it isn't helpful for my site as a whole to have those power pages drop.

I will tell you that since I have been vigorously unwinding these links the pages have continually climbed almost everyday. They have not shot back to there page 1 positions but they are moving in the right direction.

As I step back and look at how Google could look at this, I can't really say there is much use in hyper-linking with in text to a page that is already 1 click away for my user already. I could see where this could be looked at as a bit spammy or trying to game the system. I don't know if there is a threshold that once you cross you get hit with an automated filter or not. I didn't want to take any chances of being on the fence so I have moved towards removing them all. Hope it helps!

bferriolo,

Thanks for sharing that. So you're thinking you're being penalized for offering multiple ways to get to the same page on your site?

It's funny if that's so, because as a user, I like to be able to be lazy and not hunt down links, but go when the impulse takes me. I like in-text navigation and navigation bars both.

When it gets frustrating is if the site in question is a small one, and the links within a site are duplicated so much it gives the illusion that it's bigger than it is. Though even with small sites, I don't find it as frustrating if the page is clearly labeled Uses for Red Widgets each time.

No I think your fine if you have variation - I think your fine if you take the user down a path of more information via a contextual link. I think the problem ocurrs when your anchor text, main nav text, possibly your alt text as well as the text of your inbound links is exactly the same. For example: all of the above say or link to the same page for the same exact keyword "used widgets" in all instances or a combo of a couple of the examples.

The first thing I thought of when I read the cascading menu story was the fact that adding any links to a page at all, even navigation links, is playing with the on-page factors for that page. People more often consider that they're changing the the off-page factors of any target pages involved. When you get into those hover menu systems, it's easy to add 150 new links or even quite a few more (like a thousand!) without realizing it, severely changing the internal PR circulation.

Which is not to say we "shouldn't" ever add such navigation. But the average visitor will probably not use a massively expanded navigation system very much. So, if I feel there is a true user need, then I give it a try - but I proceed with care and forethought.

@bferriolo:

OK, I see now, thank you for clarifications.

However, I'm not sure why the contextual links were the ones you ended up removing. Keeping tedster's warning about extended navigation menus in mind, it might have just been that the link juice (terrible term but I'm hesitant to call it PR) from the linking pages has just been diluted by a significant increase in the overall number of links from the page.
Besides (and I hope I grasped the timing of events correctly), your contextual links predate the cascading navigation, so the new navigation would have to raise the flag first, no?

I guess, the devil is in the details: how many new links does the new navigation have? How significant is that number compared to the number of other links on the page.

yes - I do agree - everything should be in moderation. As well as if it is not good for "YOUR USERS" then it's probably not a good thing to do. Let me add - lets isolate this for now to 1 link in my main nav and me adding the 5 sub categories of taht link in the cascading nav. Of the 5 additions only 1 had contextual links in the site, because it was a good seller lets call it. Again, once added to the main nav with the 4 others, just that page got hit. Not the site, not the other inks. Just that page that was now in my main nav and was internally linked contextually. The other 4 stayed their course as well as improved in rankings.

In all cases of everything that was added to the cascading menu, the "ONLY" ones that dropped in ranking were the ones that had the internal contextual links. All other additions either maintained or got better. As well as if you did a site:mysite-com used widgets command a newer page that was added to the cascade was showing a couple of pages ahead of the page that got hit. The new page that was shoing in its place was close in relevancy but not as close to the hit page. Also, the new page was months old vs years, no inbound links vs 10 solid links.

Page content especially impacts 'theming', which then significantly impacts the value of internal linking in the website when used contextually in rich content (as well as I suspect navigation links, but to a much lesser degree)

@CainIV:

Do you mean that there are themes that have more "weight" given to links on the "themed" page? Or is it type of the site (e-commerce,informational,blog etc.) rather than theme of the site ("widgets" links are heavier than "gadgets") that matters most?

My competitor has about 6 domains for "pink widget" ranking. He has virtually no links other than spam forum account profile links and his sites are all pr0 (not grey - pr0).

The serp for pink widget is

his site1
his site1 subpage
shopping
his site2
his site3
my site (pr3 just for interest)
his site4
his site5
another persons site
his site6

Given that his offpage is totally weak - I can only surmise that his site structure and onpage content is superb.

I have printed off his source code for all his domains and also printed out word densities and common words etc for all of his domains (all his ranking pages bar the top double listed sub-page are his index pages). This seems a good way to start.

I would antisipate spotting some pattern here across his sites with regard to his onpage factors.

My problem is - that whilst I think onpage is essential, I am at the point where even though I replicate the quality and content of his work (after all - each of his sites are a replica of his others as they are all "pink widget" ranking) my pages dont appear to be moving - this makes me think that its pointless doing seo after google has FIRST INDEXED a site. Or at best - google will take an AGE to update your rank after you make onpage changes which "should" boost your rank.

I'm born and raised in America and I had the same problem trying to figure out exactly what he was trying to say. It's his experiment, he needs to be precise with his words and make sure he's conveying exactly what he did. How else can we have a meaningful discussion if we all can't be certain exactly what was done in his experiment?

Maybe he should have said "I added links on about 300 websites pointing to the nursing website. These links used anchor text that was related to winning the lottery and had nothing to do about nursing."

@CainIV:

Do you mean that there are themes that have more "weight" given to links on the "themed" page? Or is it type of the site (e-commerce,informational,blog etc.) rather than theme of the site ("widgets" links are heavier than "gadgets") that matters most?

What I was referring to was the weight given to a related page on site a linking to another page on site a.

The closer the two pages are in topical relevance, the more weight I believe is given to the specific keyword rich anchor text link. This is why I believe that on page factors do matter - when put into consideration of internal linking.

off-topic
thanks to misty and dazzlindonna

I agree with you, it has been stated in the past that an external link from an irrelevant site does not count as positively as a external link from a extremely relevant site. So why wouldn't this be true for your internal linking.

If you have an internal link that is on PAGE A that is about lawnmowers and it links to PAGE B that is about Viagra I'm pretty sure that a filter will discount that link as not relevant. I would think this would be one of the easiest things for Google to score and rank.