The Links Page is Dead, Long Live the Links Page!

32 comments

It's an oldie, but a goodie. Jim looks at good and bad links pages and what engines are almost certainly doing (and have been doing for a long time now) to discount obvious links page.

Quote:
.I’m betting that search engines hate pages like the above because it’s just so freaking obvious what the intent of these pages are…..so let’s see, If I were an engine, how would I identify these types of page so I can chuck em out of my index so as to not count these links….hum…we’ll any page with the word "link" in the url or the title tag will knock out 95% of them….I’d agree, it might throw out some "real pages" too…but for the sake of cleaning up, those babbies will be tossed with the bathwater….

There's more of course, but for many of the regular TW crowd it's obvious stuff - a good post nonetheless and perhaps a chance for us to talk about linking strategies in general?

Personally, i prefer the presell pages approach. In these days of blogs and CMS's it's so simple to put up a page of text given to you by someone with their prefered anchor text etc in the actual copy.

Comments

Blog post link value

It is a great way of getting link value, but maybe not that easy to sell and buy without Google´s editors spotting it and devaluating the whole blog mabye?

In these days of blogs and

In these days of blogs and CMS's it's so simple to put up a page of text given to you by someone with their prefered anchor text etc in the actual copy.

Of course trying to convince some people lower down the food chain, that now they have write a whole page instead of the 2 sentences they did before ain't always so easy. But I agree Jim's page is top notch.

It's Easter Time :)

maybe not that easy to sell and buy without Google´s editors spotting it and devaluating the whole blog

weather they are getting paid or just hooking up friends almost any good blog has at least a few easter eggs in it.

Looking for 'link' in the URL

?I don’t think for a second that URLs with ‘link’ in them will be what the Search engines are looking for to discount links from ‘links’ pages.

Do you know how many CMS’s use the word ‘link’ in their URLS? I don’t either, but it is a LOT of them, So a lot of people using CMS's would be at risk of losing their search engine ranking, if the search engines were stupid enough to go after the word ‘link(s)’ in the URL’s. (and I don’t think the SE’s are quite that stupid.)

I think, that if I were a smart, good looking, search engineer, I would have some time ago started looking at the onpage text (semantics) and the ratio between text and links on a page and the number of links on a page before I started going after the structure of the URL’s...

Think about it, what is it about most links pages that the search engines don’t like; Maybe it is the fact that there is little or no onpage text and only a long list of links with keyword anchors...

Besides this is all moot anyway, like Nick already pointed out, google has already figured out how to discount (or reduce the ‘importance’ of) most links pages in the ALGO some time ago.

What lots0 said

"Link(s)" as part of URLs make no problems. That's the only point where I disagree with Jim's great reminder.

Not just in the URL

I agree that a search engine would not just devalue links or kick out a page just for the words "Links, Link" in the URL.

But if they had Meta data, Title tag and page copy that reinforced the links page theme, along with links in the URL then that would be enough!

I'm not sure. Were I a

I'm not sure. Were I a Google scientist, it would be very unlikely I would care about the word "links" in a URL. However, were a marketing guy to bring me a list of 30,000 URLs that contain "links" in the title, and show me that a spam analysis showed 92% were spammy by the current internal definition of "spammy", I'd probably agree that hitting them with a filter would improve search results.

Now if "managing link farms" were to get a priority on my Google-scientist-to-do-list, for whatever reason, the first thing I would do is process those 30,000 URLs (known spam) against some set evaluations to see if they showed unique behavior (as a set) vs. some normal non-spammy sets.

I guess my point is there are different actions being taken, and any SEO analysis needs to consider the perspective of the action taker. A links page may be hit by a filter yet safe from an algo test,or vice versa. No meaningful argument there.

Jim's idea to embed links with content and maybe hide from blatant "links" filters goes part way. For the rest of the way, notice his links are all outdoor gardening links. The thematic context matches the website. If the context is thick enough, he may avoid an algo hit.

But it's not by avoiding the word links or placing the links in-line as much as it is making the links meaningful and surroundng them with sufficient on-theme semantic content to pass a likely algo or filter review.

In other words: make your site look like the rest of the web to avoid filters and algos. As Nick says, pre-sell pages.

So, if I enter the sausage

So, if I enter the sausage business and I want to sell my sausage links on a dedicated product page, I'm screwed!?! :)

FYI

Links in the URL

It's not that pages with the word "links" in the url are "banned", it's just that those pages might not pass link value.
so yes, in my humble openion, a page like yummy-sausages.com/links.htm would not pass link value.

I certainly could be wrong on that one point of my post...but I'd still say that 98% of pages with the word "links" in the url are crappy link trading pages and they will not pass link value...yes, there's exceptions, and I'm sure there are the rare few exceptions of a page with the word "links" in the URL that are "real" pages...but they are the exception, not the rule.

The word "links" in the url of that post ties in with my views from this post about link requests.

This isn't new, is it?

We had a [url=http://www.highrankings.com/forum/index.php?showtopic=6202&hl=seo-resources\.htm&st=0]thread at HR[/url] about this in May of last year, and that discussion didn't end with any real consensus.

links.htm

I've seen threads for some time suggesting "links.html" pages being devalued. Not banned, simply not parsing. There's a big difference.

..

Jimsthoughts said;

but I'd still say that 98% of pages with the word "links" in the url are crappy link trading pages and they will not pass link value...

Well Jimsthoughts you did not look at that search I provided very well, that search refutes your claim, rather obviously. If you were to look you would see that in fact it is just the opposite of what you claim, the majority of the sites that use “links” in the URL are good sites.

THERE IS NO evidence at all to show that pages with “links” in the URL do not pass the link popularity.

There is quite a bit of evidence to show that pages with “links” in the URL DO pass link popularity. For one, URLs with “links” in them are showing up in backlink searches in google.

If anyone were to check some very highly competitive keywords they might just see “links” in the URLs of pages that are ranking in the top 10 and that links from these pages still carry a lot of weight.

www.w3.org/TR/REC-html40/struct/links.html
www.nova.edu/Inter-Links/
www.uscourts.gov/links.html
www.php.net/links.php
www.bsc-eoc.org/links/
searchenginewatch.com/links/
iteslj.org/links/

I pulled these links off the first two pages of
This Search. If you have a second try running some backlink analysis on those URLs and you will see that these pages are in fact, passing link popularity to the pages they link out too.

End of argument...

I have a dream... One day, I see an SEO World that is not filled with wild guesses and supposition, but an SEO World where people check facts, run tests and review data to determine how the search engines work...

>>End of argument... Not

>>End of argument...

Not necessarily :)

What if they're showing up in GOOG becuase G knows they exist but is not counting them for ranking purposes?

..

Run the backlink analysis... that will end the argument ;-) .

>>>Added
Be sure to run the analysis on the pages that these pages (the pages with "links" in the URL) link out too.

For one, URLs with

For one, URLs with “links” in them are showing up in backlink searches in google.

Buy a link on dailycal.org - it'll show up on a backlinks check on Google, but you won't get link benefits from it.

All your search shows is that Google knows about links - it doesn't state anything about how Google actually weights the links on any of those pages.

I think Jim would have been better narrowing his statement to devaluation of links on pages named "links.html" (etc) rather than "links" in the URL in general.

And it would hardly be a far stretch for Google to exclude .gov and .edu domains from links.htm devaluation - presuming it exists.

Overall, arguing the details is missing the point with a system as dynamic as Google - but Jim's overall assertion about naturalising link structure to "camouflage" links from devaluation is worth paying attention to.

testing

One day, I see an SEO World that is not filled with wild guesses and supposition, but an SEO World where people check facts, run tests and review data to determine how the search engines work

I've been doing some testing trying to isolate things as much as possible to find out what the alogo's like and what they don't like. MSN is the least evolved at this stage and Google is the most advanced (ack, I said something good about google!). The thing I'm discovering is for Google it's rarely one factor that you can isolate that makes something good or bad, but put things in combinations and all sorts of good/bad magic starts to happen.

Buy a link on

Buy a link on dailycal.org - it'll show up on a backlinks check on Google, but you won't get link benefits from it.

Do you have any evidence at all to show that no link pop is being passed ONLY because there is the word 'links' in the URL? Are you sure it does not have to do with page structure or the ratio of links compared to the onpage text? Or maybe no link pop is passed cuz the dailycal.org runs all its links thru a php database? I would vote for the latter as the reason no link pop is passed...

And no, just showing up in the google IBLs is not proof or even strong evidence, but... I have yet to see even the most minute evidence that even remotely points to links on these pages not carrying any weight, based solely on the fact that 'links' appears in the URL.

...for Google it's rarely

...for Google it's rarely one factor that you can isolate that makes something good or bad, but put things in combinations and all sorts of good/bad magic starts to happen.

BINGO! We have a WINNER!

Or maybe no link pop is

Or maybe no link pop is passed cuz the dailycal.org runs all its links thru a php database?

Not an issue - point is, you used as "proof" that links were counted by the fact that such pages show up in a backlinks check.

I simply pointed out that just because Google can spider links on a page and trace a relationship to other sites, doesn't mean to say those links will pass on benefits.

you used as "proof" that

you used as "proof" that links were counted by the fact that such pages show up in a backlinks check.

Nope, what I said was;

There is quite a bit of evidence to show that pages with “links” in the URL DO pass link popularity. For one, URLs with “links” in them are showing up in backlink searches in google.

And

And no, just showing up in the google IBLs is not proof or even strong evidence...

I NEVER said that showing up in a backlink search was "proof" of anything, I believe I said it was "evidence" that is quite different than "proof".

No proof

URLs appearing in BL searches don't necessarily pass PR or topic relevancy via anchor text. However, most do, even links pages which aren't that well optimized, future-safe and content-rich as Jim's example. Sometimes we think everybody is as 'evil' as we think we are in the eyes of the engines ;), but there are still lots of non-optimized 'honest' links pages out there, and many of them are not devaluated. Having 'links' in the URL, title tag, and on the page may be a risk, but there is no evidence that it is widely used as an indicator for link spam.

My bad - "evidence" and

My bad - "evidence" and "proof". Seems like we're going to tumble around a semantics issue here. Point was underlined that backlinks checks will show links which do not parse PageRank or anchor text benefits, which is something worth noting.

Graywolf's Point Stands

Quote:
The thing I'm discovering is for Google it's rarely one factor that you can isolate that makes something good or bad, but put things in combinations and all sorts of good/bad magic starts to happen.

I don't know of any way to test the effect of this one element without some other element(s) skewing the results, so I don't think we're going to come to any agreement on its effect.

.

Seems like we're going to tumble around a semantics issue here.

Huh? You said, I said something, that I did not, I corrected you, simple as that.

Point was underlined that backlinks checks will show links which do not parse PageRank or anchor text benefits, which is something worth noting.

Again - Huh? Would you care to expand on this? For example, just how did you determine that the links are not parsed(3a and 4) by google?

And... what about ALL the pages with 'links' in the URL that do show google pagerank? Should we just discount them because they do not fit with this theory that the word 'links' in the URL discounts a page's ability to pass on link popularity?

"Well Jimsthoughts you did

"Well Jimsthoughts you did not look at that search I provided very well." (allinurl: links)

Few comments, 1. I did, 2. Just because google shows that there are pages indexed with the word "links" in the url does not mean that these pages pass any value. A search for "allinurl" just means that those pages are indexed.

Even if you were checking a sites backlinks and it showed a backlink from a page that had "links" in the url I still wouldn't put any value on it. Google doesn't show lots of things that "count" and shows many things that probably "don't count".

All debating aside, I'd say "if you have a choice of a page name, try not to name it "links"" ....just in case....I tend to lots of things "just in case".

Just because google shows

Just because google shows that there are pages indexed with the word "links" in the url does not mean that these pages pass any value.

I agree. I only asked if you looked at those results after your statement that 98% of all sites that use 'links' in the URL are 'crappy link trading pages'. Cuz if you looked at those results you would see that that statement of yours is not correct.

You know there is no need for disagreement or debate, just run the backlink analysis on those results yourself.

I tend to lots of things "just in case".

Reminds me of an old neighbor of mine. He used to wear a steel combat helmet all the time. When I asked him why, he told me it was just in case a jet passenger plane dropped it's waste holding tank while flying over him... ;-)

Steel combat helmet ;)

Maybe I do wear a steel combat helmet, but personally, I'd never deal in any link trades with any "links" named page.
Oh well, I can agree to disagree on this one :)
Cheers!

Again - Huh? Would you care

Again - Huh? Would you care to expand on this? For example, just how did you determine that the links are not parsed(3a and 4) by google?

Now you're confusing what I said, which was as you quoted:

Point was underlined that backlinks checks will show links which do not parse PageRank or anchor text benefits, which is something worth noting.

You earlier stated that link pages being listed in Google was evidence that the links were counted:

There is quite a bit of evidence to show that pages with “links” in the URL DO pass link popularity. For one, URLs with “links” in them are showing up in backlink searches in google.

I simply pointed out that being listed in backlinks is no indication of the weighting of those links.

.

Brian if we are going to debate, your going to have to stop going around in circles. Your attributing quotes to me where I was quoting someone else.

Brian you said, the URLs with the word 'links' in them were not being 'parsed', I asked you to expand because your statement made no sense.

All pages are 'parsed' when googlebot indexes them, so if a page is in the google index it has been 'parsed' and those pages are obviously in the index so they MUST have been 'parsed'.

Brian have you done a backlink analysis of those URL's that have the word 'link' in them? Cuz if you had you would be able to see for yourself that, all things being equal, these links do in fact pass that link love on.

Parsed

Sorry lots0, I think you have mine and Jim's comments confused together - I don't believe I stated that "URLs with the word 'links' in them were not being 'parsed'". I mentioned about PageRank not being parsed from devalued pages, which is a different issue.

Of course our link-weighting algorithms are the first line of defense, but it’s difficult to catch every problem case in adversarial information retrieval, so we also look for problems and leaks in different semi-automatic ways. Reputable sites that sell links won’t have their search engine rankings or PageRank penalized–a search for [daily cal] would still return dailycal.org. However, link-selling sites can lose their ability to give reputation (e.g. PageRank and anchortext).

Thanks Brian

Brian, you're quote from Matt hits the nail on the head here.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.