next »

[patch]

Guys,

I'll kick things off then.

I know very little about SEO except:

• maybe using keyword rich domain names will help with ranking
• keyword rich urls e.g. www.bluewidgets.com/bestbluewidgets.php as opposed to www.bestbluewidgets.php?prod=12345
• I know nothing about the optimal internal linking structure or how important that is
• I don't know whether it's best to use subdomains or not
• and on and on it goes ...

So, I s'pose the question is ... where do I start?

Cheers.

[vsloathe]

SEOMoz just released their list of "ranking factors" yesterday.

http://www.seomoz.org/blog/ranking-factors-version-3-released

I would start by reading a lot.

[Bompa]

Guys,

I'll kick things off then.

I know very little about SEO except:

• maybe using keyword rich domain names will help with ranking
• keyword rich urls e.g. www.bluewidgets.com/bestbluewidgets.php as opposed to www.bestbluewidgets.php?prod=12345
• I know nothing about the optimal internal linking structure or how important that is
• I don't know whether it's best to use subdomains or not
• and on and on it goes ...

So, I s'pose the question is ... where do I start?

Cheers.

Shit, no one knows that stuff for sure.

I do know that keywords in the Title moves my site up, remove them and the site goes down,
and that happens in any niche that I have entered.  The same is true for keywords in the body of the page.

I can not do that experiment with domain names and URLs easily.

"Internal linking stucture" is something that the WH SEOs made up to keep their jobs.   

According to some old, original, PageRank papers, there are two ideal strategies for internal linking.

1.  The homepage links to ALL other pages and ALL those pages link back to the homepage.
2.  Every page links to every page.

Either way is possible with a small site, but with bigger sites, niether is possible so people just
try their best.  I think #1 tries to give all PR to the homepage and #2 spreads the PR evenly.

Bottom line is that no page is orphaned.

The rest is about well anchored links.

So, read what the gurus have to say.  Remember it's just their opinion.  Then experiment.

imo,
Bompa

[rcjordan]

>"Internal linking stucture" is something that the WH SEOs made up to keep their jobs.

Agreed. With the one exception that you can use the nav links to raise kw density, sometimes substantially.

[isthisthingon]

@perkiset is an SEO    and many others on the Cache are deeply involved with it too.  I'd help if I could but you're definitely in the right place.  Goodluck

So, read what the gurus have to say.  Remember it's just their opinion.  Then experiment.

Words to live by...

[perkiset]

WhiteHat Simple page-based optimization:

After having seen all manner of things become significant, then trivial, then significant, then trivial ad infinitum it seems, my personal philosophy is to do every single trick I know on every single page every single time. To me, SEO is about discipline much more than the technique du jour.

Things I always make sure of, regardless of it they're significant right now:

• keyword rich URLs that use dashes to separate words (not underscores) - there is much debate around this issue, and several of the engines now see underscores as the same, but dashes are by default more accepted as word separators
• A lush header: description, keywords, title, favicon. A DOCTYPE. Well formed, conforming HTML. Seems silly, and again, lots of debate over the effectiveness, but it'll never hurt you to make this part of your deployment protocol.
• No tables, or anything structural to your HTML. Use DIVs, SPANs etc to structure your page. Tables, particularly are a killer because the engines have a harder time developing a sense of context or semantics around your page if content is broken up into cells. The more you can approach making your content a single, unobstructed chunk of text near the top of the file the better. I know I'll even get crap about "at the top of the file" but it *used* to be important, it may again someday. One more time: if you do this with discipline, then it'll never screw you later.
• Use H tags. They are important. I'll say no more here.
• Push JS and CSS offpage if you can swing it. The smaller and tighter your HTML the better. The less "other stuff" the better.
• Use every opportunity to put your keywords on the page, within what you'd consider to be "natural looking" as you can. In other words, never miss the opportunity to supply an ALT tag for an image. Make the URL for images as descriptive as you can. Make sure that, although you want your text readable, you talk WITH your keywords rather than around them, even though around them might make for better reading.
• Use links judiciously. In my case, I only put <a href into places where I KNOW I want to pass juice. Other than that, the links are obscured or JSd. I am not one that thinks you should horde juice. I have personally seen evidence that when a site as a certain component of being "part of the information continuum" rather than a link juice bogart, it does better. In other words, pick some places to link to that corroborate the story you're trying to tell and bleed out some link juice on them.

There are more, but that's a good start for now. Bomps post re. internal linking is strong and VS' point is essential: read read read read read. Then try. Lots. Read every word that RCJordan puts out there carefully ... it is based on a great many years of experience.

Good luck!

[perkiset]

@perkiset is an SEO 

Most kind ITTO. But the truth is, I built The Cache so as to attract the likes of you, Bomps, RC, VS, Nutter etc and become smarter.

Thanks tho

[isthisthingon]

Most kind ITTO. But the truth is, I built The Cache so as to attract the likes of you, Bomps, RC, VS, Nutter etc and become smarter.

Brilliant plan. You should have been a genius   But when it comes to SEO I'm probably as helpful as an SEO for Dummies book - abridged... written in Coptic.

[deregular]

Another good tip when trying to figure out what will work..

Put yourself in the search engine's shoes.
Not only what you would do in order to spit out relevent results,
but what you would do to spit out relevent results on a massive scale.

Some things simply can or cannot be used to gauge a sites ranking
because it may harm specific legitimate niche rankings or benefit others too much.

One example, they simply cannot persecute for somebody elses actions.

Another, keyword density while it was a factor early on, not so much now,
even then they could not penalize for upto a certain density because by nature
online shopping sites as well as others have a very high keyword density due to
their directory type structures.

More of a mindset tip, but hopefully useful nonetheless.

[patch]

Well, thanks for all replying in such a detailed manner.

That's good stuff for me, particularly like the 'checklist' from perkiset, I'll definitely work to that... just to clarify though, javascript & css in external files rather than embedded, that right?
and the debunking of the 'internal linking structure' stuff too ...

Got a few more questions though, I'm not asking to be spoon-fed but can you be more specific about what/who to 'read, read, read' (thanks vsloathe for the link, reading that stuff now) - left to my own devices I could easily spend endless hours reading the 'wrong' advice so any pointers there gratefully rcvd.

Agreed. With the one exception that you can use the nav links to raise kw density, sometimes substantially.

with regard to the above, I read elsewhere that it would be of use SEO-wise to potentially 'update' nav links and page titles on the site based on incoming keywords, would that be a reasonable way to proceed?

My other questions could well be answered by finding out what to read so I'll save them for now.

Cheers

[rcjordan]

>Read every word that RCJordan puts out there carefully ... it is based on a great many years of experience.

That's bullshit, Perk, but thanks anyway.  I have experience, yes, but it's outdated.

My own philosophy re (long-term) seo is well-said here by others; plan based upon the action you believe the se's will be forced to take in the future.

>potentially 'update' nav links and page titles on the site based on incoming keywords, would that be a reasonable way to proceed?

Dynamically? On-the-fly? As in, alter the landing page to be whatever the bot wants?  Oh, that'd be black.  It used to work well, probably still does, though I haven't seen as much of it of late.

Or did you mean manually? Sure.

[perkiset]

... just to clarify though, javascript & css in external files rather than embedded, that right?

For WH I'd advise it all the time, although I think there are some that would debate me. IME, it works nicely. I don't see any benefit to having it *in* a page (at the moment anyway, there is no evidence that classnames, ids, styles or JS variables etc could be used for keyword stuffing) so I'd find it good practice to push it out in any case. Makes for better coding practice and separation of content from appearance.

Got a few more questions though, I'm not asking to be spoon-fed but can you be more specific about what/who to 'read, read, read' (thanks vsloathe for the link, reading that stuff now) - left to my own devices I could easily spend endless hours reading the 'wrong' advice so any pointers there gratefully rcvd.

Take a couple sites that are beating you up in the SERPs. Read the source. Then go find research/evidence/discussion about elements that you see in the source and why it works/doesn't work etc etc. You will begin to see patterns and know how to ask questions, and where better to read then. But start here and the Synkd8. Search for "Keyword Stuffing" over there and read all the threads you find, for example. Then go see if you can find examples of that usage. There's out there. It's not easy, but it's not that tough.

with regard to the above, I read elsewhere that it would be of use SEO-wise to potentially 'update' nav links and page titles on the site based on incoming keywords, would that be a reasonable way to proceed?

You mean dynamically changing the page to reflect the inbound keywords? Well, you've kind of crossed over a couple points there. If you're talking about doing something SEO wise, then you're talking about appealing to the spiders: in which case there'd be no inbound keywords. Inbound keywords would only be found as the result of a search. So changing keywords in a navlink based on inbound phrases is pointless. May very well be good for conversion though - customizing your page to appeal more directly to the inbound keywords is a time-honored technique and can be very effective. But changing your pages for the spiders is cloaking, or at least IP Delivery, and it's just a little more difficult to make a case for WH IP Delivery anymore, although there are those that do it   this bends towards mosaic cloaking, which is an advanced technique and probably beyond the scope of your question here. I have a a thread on it here: http://www.perkiset.org/forum/css_html_seo_cloaking/mosaic_cloaking_an_explanation-t657.0.html

[perkiset]

That's bullshit, Perk, but thanks anyway.  I have experience, yes, but it's outdated.

Liar.

[rcjordan]

>Inbound keywords would only be found as the result of a search. So changing keywords in a navlink based on inbound phrases is pointless.

Say you have 'hotels' buried in the body of the page and find that you have a trickle of traffic coming from deep in the serps. Putting the kw in the navlinks might add something ...but based on what I've tried recently, it's not proving to be worth the effort.

[perkiset]

Well yeah... not dynamically, but adding the keywords statically to the pages so that the spiders see them, that *may* be of benefit. But I read him as dynamically changing nav links based on incoming keywords ... which meant changing the page for surfers based on how they got to the page. Don't see that as a benefit to SEO at all.