Keyword Linking Ranks Wrong Pages

[timemachined]

Two pages

'blue widget hotels' and 'red widget hotels'

A text link on 'red widget hotels' page is 'blue widget hotels' linking to 'blue widget hotels' page

What ranks for 'blue widget hotels' page? Despite 'blue widget hotels' being optimised.

Well obviously, 'red widget hotels' ranks for 'blue widget hotels' because of the key phrase link on the page.

Common sense... a frustrating and stupid search engine. Giving emphasis not to the link inbound to the relevant page, but to the page the key phrase link is on.

It's been like this for ages but I can't figure out why G does it. It only serves to rank the wrong page for a term instead of the correct page. Furthermore the correct page if I resubmit will sit page one, the incorrect page that G chooses suppresses ranking and will sit 3rd or 4th (of ten results per page).

Recommendation

Do not keyword phrase link to other pages
Do the silly thing you've seen others do, and keyword phrase link the page to itself.

[keyplyr]

“Ours is not to wonder why. Ours is just to do or die.” - Alfred Tennyson

I think the keyword linking would perform as you intend if there was enough additional content and specific keywords/phrases supporting it.

[timemachined]

If I can figure out that G is doing that to me, you can be confident that I have optimised the page for keyword strings. G is amazing at ranking unrelated pages just because it has a linked key phrase on it. I had to remove related articles because of this continuing problem.

[keyplyr]

Not that it isn't happening, I've just not seen it so I can't point to anything suspect to experiment with.

Possibly because all the pages (my own & my clients) have all been in existence for quite some time and this anomaly could be something happening to newer pages getting indexed or some other yet-to-be-known factor triggering this affect.

I've always taken careful steps for my pages (all similar theme) to not compete against one another. I was especially concerned about the situation you describe when 4 years ago I added interpage text linking similar to Wikipedia, but it went (and stayed) without negative issues.

My pages aren't ranking for the out going text links (as you describe) possibly because I have a half dozen or more other text links going to different pages using different keywords... mixing it up.

At any rate, if this was happening to my pages I would take a couple lesser pages and do A/B testing until I hopefully stumbled on the right combination.

You could also throw away one of your pages (serving a 410 Gone, no 301) and bring back that content on a newly created page with a new URL and Title and wait and see if the same thing happens. Just an idea.

[masterjoe]

I am getting really irritated with the way Google is handling this too -- I have even deoptimized a page (which wasn't even really "optimized" in the first place), because it is cannibalizing a far more relevant page. The page it insists on ranking is an LSI variation of the term that is actually getting searched. It has a few links, and one is from a website that has the keyword I am trying to rank, which is why I'm assuming it does this.

Its quite sad that we must compete against our own websites, whereas the big boys like Amazon freely get multiple rankings in the results anyway.

[tangor]

When I've come across this, and it isn't too frequent, I mix the content of the "ranking page" into the "relevant page" I really want to rank and it works a treat. User gets what they are looking for and I get the eyeballs on what I want them to see.

Sometimes we can get too granular and out fox ourselves.

After all, it is the user we have to please and they are the ones putting that search into g or b or y or whatever.... Code for them, not the SE

[martinibuster]

1. Anchor text within a web page is said to be a signal of what that web page is about.

2. It's possible your page is not structured correctly to communicate what it is about. Run your code through the W3C Validator and tick the box for the OUTLINE to visualize your page hierarchy.

[Williamhollingworth]

Adobe example on ranking for "Click here" is pretty interesting :-) although mainly thanks to external links and anchors

[martinibuster]

That's an oldie! Used to be that way. :)

The anti-Google bombing tweak (or maybe something else) has changed that SERP.

Back on Topic: Internal Anchor Text

It's a good reminder that heading tags and title tags aren't the only signals of what a web page is about...

[keyplyr]

It's a good reminder that heading tags and title tags aren't the only signals of what a web page is about...

Exactly, and with AI contributing "intent" to the mix, it's even more important to support page topics with strong content.

[Robert Charlton]

Structuring your site so that pages with general content link to pages that contain more specific content can help prevent this.

[samwest]

I was looking through my anchor text report in WMT and it is loaded with garbage. Foreign phrases, nonsense phrases and only a tiny fraction of the report is legitimate anchor texts that relate to the site. I'm not sure how accurate that report is or how much weight these metrics carry with Google, but if it's used in any way to rank my site, it's no wonder why the site has continued to sink over the past 10 years.

I am guessing most of this comes from those garbage scraper sites that mass linked to me over the years. The site dates back to 2000. The first 10 years were great with steadily improving rank and traffic, but I'm guessing once this negative SEO kicked in, presumably from hackers and those who want to see me fail, it's been downhill ever since.

Sadly there seems to be little to nothing you can do about. I did a disavow on all the spamy / scammy looking URL's, many which were #*$! related, Russian or were from .xyz domains. Regardless, this has not erased any of the garbage. I posted to the WMT help forum, but it produced no help...one member just guessed it was no longer relevant and to ignore it and instead focus on content and "answering the query". I have been...it doesn't help.

[Rlilly]

martinibuster + is correct, the proof is Trump: he is more about Twitter based on Anchor Text in links than being President..

[martinibuster]

Not really. That's not what I meant. :P

1. This discussion is about INTERNAL anchor text.

2. Google shows searchers what they want to see. When a certain percentage of people search on that persons name they want to see a Twitter account.

3. Anchor text as a ranking factor will not outrank what a user wants to see.

Please let's stay on topic. This is about Internal Anchor Text.

Thanks,

Roger

:)

[smilie]

@ OP: yes they can randomly confuse it.

They weight 200+ indicators. And modify them monthly to confuse us and spammers.

So your particular page may have a specific scenario whereby a different page for the same KW ranks above it. Happens all the time.

On our ecomm G randomly selects specific pages that algo thinks are better for a "green widgets" vs. "blue widgets", and we just play along and let them rank rather than try and change it.

Could be anything, your page can have over-optimization penalty, sounds like it from this: "Despite 'blue widget hotels' being optimised". It this is the case, total negative weights of certain page factors outweight positives, so it sinks below the next available page, "green widgets hotels".

[samwest]

@martini - did not notice specific reference to internal links in the OP, but it's not like we haven't wandered off topic before.
My problem seems to be with external anchor garbage links to my site, but we all know that is virtually impossible to fix.

So back on topic, if on my home page I link to a sub page using the term "blue widgets" to a page ABOUT blue widgets, that is bad?
I do it all the time, but I am also swirling down the drain fast. I figured it was logical from a user stand point.

So what are we supposed to do, anchor text a link on the blue widget page using the term "blue widgets" to it's own URL? That's wack and IMHO a bad UE. That's what I got from the OP's last two lines...that or I'm missing the boat again.

[smilie]

@samwest, G is tightening the on-page factors , calling it "over-optimization penalty", essentially.

If there's too many "blue widgets" links and KWs on the page, and "blue widgets" is a money KW (G knows which ones are), you are going to sink very fast.

All I do these days , after 5 other daily jobs, other than making pages look better for the users, is figure out how to remove over-optimization penalties on older pages. Takes moths to get results if any, and sometimes doesn't even ever un-Panda these pages.

Look at this topic I created earlier, "Does Matching Title & H1 Trigger Google Overoptimization Penalty? "
[webmasterworld.com...]

[30K_a_month]

over-optimization penalty ?

One of my good ranking pages has its keyword as H2 at least 10 times

Bolded at least 10 times

Another 10 times in the text

and 10 images all with the keyword in the alt text and the file names

And a H1 at the bottom

1000 odd words

This page has moved between 1 - 3 serp position for the past 2 years. Med/High keyword

What over optimization ?

[samwest]

Ok, so if I have a overall page about "general widgets" and as I am discussing these widgets I decide to drill down and mention "blue widgets" or "red widgets", are you saying that if I link the phrase "blue widget" or "red widget" to a sub page that details blue widgets or red widgets, that this is considered an over optimization?

It is only intended to help the user better understand "widget specifics"...or does Google prefer websites with one sprawling page about all widgets. That would seem to be a very poor UE to me.

Now, on another thought...are you really sure what you are doing is actually fixing the problem? You mention that is can take months and sometimes never even works. This begs a few questions:
1. Once marked and over optimizer, always an over optimizer?
Sounds like Google is citing you with a crime that will never leave your webmaster record...and...
2. Is Google simply removing the reward by removing or delaying the response to your stimuli?
For example, when a doctor hits your knee with a hammer and you react by kicking, he knows your nervous system is working properly. If I make a change to my site and my rank jumps, then we get a small glimpse into how the algo works, and I don't think ANYONE on this or any other webmaster or SEO forum would agree that this is how Google works or wants to work, If they did react, and it was repeatable, everyone would start doing it and this would usher in a new wave of web spam...that's likely why Google seems to use a type of reverse psychology or delayed reaction (sandboxing, yeah I know that doesn't exist) when we make site changes.
3. Finally, wouldn't removing these type of links constitute reverse optimization in order to improve your rank (which is really optimization) and wouldn't that be equally illicit according to Google's twisted logic? Seems anything we do with the goal of improving rank is bad.

Not trying to be argumentative or or a smart ass, but I'm just trying to better understand what you are implying or surmising.

I rather agree with 30k's comment...if you slap a sloppy mess up there, it (in too many cases) ranks better than quality, well thought out and informative content.

In my own vertical, the number 1 result for a high volume (three word) money phrase belongs to a site that has not been maintained in years, uses flash, sprawls over 1000 words, is keyword loaded and uses spammy repetition in the text, does not use any heading tags, used to participate in link farms, and uses broken SSL on their order page that pops a security warning. That's the number 1 result and has been holding for almost a year now. However, that site is at page 100 for nearly every on term in that vertical. Oh, and it also uses the blue-widget bold anchor text linking style you describe.

Please correct me if I am totally missing the point here...I probably am.

[samwest]

One more thing...would you recommend doing a no follow of internal links?
That whole link juice thing is too confusing.

asked and answered...I guess, as that was from 2013: [searchenginewatch.com ]

and a follow up to my previous rant...regarding internal anchor text, from the horses mouth: [youtube.com ]

[smilie]

@30K_a_month : what over optimization?

Life is different from every angle. Yet it happens all the time. It also doesn't happen on strange occasions as well.

From that description, your KW must be over 5% of all other content. If that's the case, other strong factors may contribute for it to hold such as strong pattern external and/or internal links. Or site being white listed by G. Or great site that's always thumbed up by thousands of SERPs checkers who work for SEs. For a small bus site, over 5-7% is, generally, doom by Panda.

>> @samwest: 1. Once marked and over optimizer, always an over optimizer?

Not necessarily. I was able to pull some pages significantly up in SERPs.

>> @samwest: Seems anything we do with the goal of improving rank is bad.

They certainly try to create lots of FAD and traps.

[JS_Harris]

This a classic example of why internal link structure still matters, more so than many webmasters realize.

Most sites have a typical link structure with sitewide navigation links, categories and footer links. On this type of link structure it's the index page, followed by the category pages, which are seemingly most important. Even if category pages rarely rank you are still sending the message they are most important with this link structure.

You can cut down on what the opening post describes by making individual articles appear more important and category pages appear less important within your internal link structure. To do this you can simply more heavily interlink articles with each other to increase their perceived value within your internal link structure. You could take it further by not having those category links be sitewide, or by only linking to the category the article is stored in unless on the home page etc, but just interlinking articles with related articles, preferably from within the content area, is enough.

Once individual pages are deemed more important than category pages you will see this happen less. You see it happen less on older pages because as those acquire links from other sources they become interlinked, and thus more important, naturally.

Note: I think Google does a good job of eventually correcting this issue but the first pass initial ranking system does not catch it all the time, you need to wait for subsequent deeper calls and, perhaps, even a human rater who will downgrade the page from "red" when it's clearly about "blue".

Hope that helps shed a little light on the issue. There may be other causes but in my experience that's the best way to protect yourself and build a better site. You might also consider condensing your pages so that one page can talk about both red and blue. The mentality of having a page per keyword or keyword phrase REALLY works against you now, fyi. Google prefers comprehensive and will send a good page traffic on a wide range of keywords.