Internal links with rich anchor text

[AlmostHuman]

We've recently decided to double our efforts as far as content goes - more blog posts, more useful information in our knowledgebase and such. All of that is going to be published on our own website.

Of course we want to get as much from it as possible, so we want to add links to relevant products we sell and services we provide in new posts and articles using rich anchor texts.

What is the current situation on internal rich anchor links at the moment?

Please, don't reply with posts like "think about how much user is gonna benefit" or something like that. The links are going to be relevant - it's our site, after all, but since we have a choice between rich anchor links and anchorless ones, we would like to lean towards the former.

[netmeg]

Actually, I had a response but as this is the second or third time you appended the "don't reply with" statement, I decided not to bother.

[JD_Toims]

The right answer is the one you don't want to hear. You're not going to outrank someone these days because you linked "green widget" and they linked "read more".

[lucy24]

choice between rich anchor links and anchorless ones

What does "anchorless" mean?

:: unavoidable mental association with plumber recommending washerless faucet, where "washerless" means that instead of a washer there's "a rubber thing" yup makes all the difference ::

[FranticFish]

As others have said, you're clearly hoping someone has an answer along the lines of 'if you have 1000 words you can insert 10 anchor text rich links'.

There obviously is an acceptable and non-acceptable ratio for this - but why are you looking to gamble all your 'doubled efforts' on some arbitrary figure from a forum post?

No-one knows the answer unless they've tested this with sites they're ready to forget once/if they get thumped by Google. And their answers won't help you anyway. If that's the game you wanna play, then don't ask, play it yourself and work it out yourself because the answer will be right for the sites you build in the market you operate in. I can't tell you where the line is drawn because I don't care to risk a site to find out where it is - and even if I did the acceptable threshold might be affected by so many other factors that it makes my head spin.

But if you want to know what definitely won't get you in trouble, then just try to
(a) link when it serves a purpose, and
(b) describe what you're linking to accurately.

[netmeg]

Oh but you're not sposed to reply with that, FF.

[Robert Charlton]

I think FF's reply is right on point. It's worth the OP's while to read it several times.

The OP should also pay careful attention to what JD_Toims said.

[Dymero]

Here's the thing. I've seen some evidence that internal linking alone can be powerful enough to boost a page relatively high in the rankings, but the effort has to scale. Posting the link in a couple blog posts just isn't going to do it and if your site is small, forget about it.

There is also probably a point of diminishing returns because, even if you can scale the internal linking effort, you can't put a link to all your products on every page, or else you'll trip some kind of over-optimization or spam filter. And even then, you need to have a fairly good external link profile, because that's all that's carrying those pages.

As for whether to target with exact match all the time, I would not unless your product is actually called the rich anchor text you'll be using. For example, when we link, we're less likely to use [BrandName Widget] as we are just [BrandName].

[JD_Toims]

I've seen some evidence that internal linking alone can be powerful enough to boost a page relatively high in the rankings, but the effort has to scale.

Uh, that may be "what it seems" but if I use wikipedia as an example, does it rank for a *huge* number of terms because they internally link key-phrases, or does it rank because it's usable, has depth of information and makes sense to visitors, even if the information is not always accurate?

WikiPedia couldn't care less about PageRank distribution -- If there's a "topic" they have more info on, it's linked. If there's a source to cite, they cite it. It doesn't matter if there are 5 or 100 or 500 links on a page, and [I haven't counted but in "reviewing sources" it appears] sometimes, the "tech babble sources" without explicit key-phrases in the citations out-number the phrases linked on a page -- Bottom line: They do what's best and makes the site the most informative/usable for the visitor, and, imo, that's why they rank, not because Google's got a love affair with Wikipedia -- It's really the other way around. Google ranks WikiPedia, because it's a site that makes sense to people and people want to find.

Sh*t! I just gave the answer I wasn't supposed to according to the OP, because rankings are becoming more and more about usability and what people want to find while becoming less and less about what *bleeping* words someone decides to link to provide a depth of information -- I'm sure it's been said before, but: "Keywords are dead, long live keywords." -- Can anyone else imagine the "blow back" if wikipedia decided to link [read more] next to every phrase/word they have linked on a page rather than linking the phrase/word itself, yet somehow, for some reason (gotta be a Google glitch) still ranked? If it wasn't a "Google glitch" it would be like all the 2005 SEO concepts went right out the window and SEO wasn't about "linking all the right words" anymore, but required building a site people actually want to find, which is much more difficult.

[Robert Charlton]

From the OP...

What is the current situation on internal rich anchor links at the moment?

It's actually not clear whether this question is asking about nav links or about inline contextual links... but I think we're assuming that the OP is thinking contextual.

What's often missed is that NYT or Wikipedia contextual links, which are usually used as "footnotes" for the user, happen to "work" also for SEO because the the New York Times or Wikipedia articles attract good inbound links.

Thus, the contextual nav links have some link juice to redistribute, and they redistribute it to naturally related topics. Keyword-rich contextual links from just any old page on any old site don't necessarily work at all... and could well create problems.

It doesn't matter if there are 5 or 100 or 500 links on a page...

I think it does, though Wikipedia can probably withstand the problem more than most sites. Note the Wikipedia internal article on the "overlink crisis" I quote in this thread. The article was written not with SEO in mind, but rather with regard to usability....

Introducing heavy internal linking to well ranking site
http://www.webmasterworld.com/google/4119868.htm [webmasterworld.com]

IMO, too many links can present too many choices for the user, and that's never good.

[JD_Toims]

I think it does, though Wikipedia can probably withstand the problem more than most sites.

In the context of PageRank flow, I definitely agree, and that's where I think most "get stuck", but in the context of "overall value", which includes "trust" I think there's a distinct difference -- If you ask me to cite a source right now I don't have them in front of me, but in "one of those articles/posts" by someone like Rusty_Brick or Bill Slawski or even a MC/GoogleGuy comment or someone else "who really studies things / knows their stuff", I remember reading about a connection between where you link and "trust" relating to your site/page -- Linking to "depth" / "a trusted source" has essentially the opposite effect of linking to "bad neighbors" or "questionable areas".

So, from the overly discussed perspective of PageRank distribution [which many seem to think is the be-all/end-all way to rank, even though the results indicate differently], then I'm with you, but from an overall ranking perspective, especially including "trust" and "not simply talking about a topic" but rather providing a "depth of information" on a topic, then I have to disagree a bit, because if you look at WikiPedia's pages, you [or any reader] swill ee if a word/phrase can be further defined, it is, and if a source of information is not cited/linked they're asking for it to be.

They don't say, "If there are already 20 sources on this page and you have another complimentary/conflicting source for the edit made or information provided, please don't post or link it...", or, "If there are already N links on this page and there's another page defining the term/phrase used, it should not be linked" -- They're definitely linking to "definitions" and "expanded information" as well as trying to get sources cited and they're not "limiting" those expanded-definitions/sources as long as the definition/source is considered valid/reliable AFAIK. If you know differently, then please, let me know.

Bottom Line: They don't care as much about the "link weight/pr distribution" as most "SEOs" do [or at lest seem to], because they're trying to be "usable and informative" rather than "SEOing for 2005 in an effort manipulate search engines to rank them the highest", imo.

[phranque]

the List of Internet top-level domains [en.wikipedia.org] wikipedia page has over a thousand links and i can usually find exactly what i'm looking for in seconds.
if you search google for "tld" or "tlds" or "top level domains", that page is owning it.
in usability terms, if they decreased the number of links on that page it would likely take me longer to find something.

[Robert Charlton]

phranque, great example and I get your point, but it's also a very different kind of page from what I was talking about, which is to say a prose "article" with contextual links used essentially as footnotes.

The "List of Internet top-level domains" is just that, a multi-column list, presented in extremely well-organized tables, with alphabetical sorting of TLDs and reference links for each TLD and geo entity represented, along with language on some tables, with actually fairly sparse linked contextual references in the main chart's notes. Yes, of course you can find things quickly... and it would be foolish to, say, split up or paginate such a page.

That in no way relates to what I was talking about, where there might be so many apparently random links that the user would be overwhelmed.

Again, that overlinking article was about usability. SEO and PageRank distribution never got mentioned. When usability is considered, PR distribution is essentially taken care of, and I took pains to note that in my comments.

That said, don't try to do at home what Wikipedia does. You may simply not have the link juice to support a thousand links on a page. IMO, prioritized PageRank distribution is still a factor in site design, as is semantic clarity in nav structure.

[Dymero]

Uh, that may be "what it seems" but if I use wikipedia as an example, does it rank for a *huge* number of terms because they internally link key-phrases, or does it rank because it's usable, has depth of information and makes sense to visitors, even if the information is not always accurate?

Probably some of column A, some of column B. It's useful information, but isn't always terribly relevant to the page it's linking from. And, as Robert points out, they have an overlinking problem. However, if you subscribe to the "brand overuse" complaints some SEOs have, it's untouchable because Wiki is a top site.

That said, Wiki is a great example of a site where the internal linking structure is strong and scales because of massive manual effort (and maybe a few Wiki bots?). True, Wiki links well externally, but it's also linked extremely well internally.

When I get some more time, I'd like to do a formal study and see if a strong internal linking structure truly can help boost a page, even if it has few external links itself, or if it just appears that way and there is something else going on.