14 replies to this topic

[niemi]

I just wondered because it seems that everybody is talking about headings, content freshness and META tags, still I see numerous sites that doesn't have any headings, content from the 90's and with no META tags having a PR of 5+.

It should be easy to start the list out with:

1: Incoming links.

[JohnMu]

1. read + watch
2. think
3. test
4. think
5. monitor results

I think I'm just old fashioned.

When it comes to a moving target, there are no static action items. Things change. Don't rely on anything to be the main item with regards to SEO (though links are a pretty good guess --- but what kind of links?).

Personally, I think the most important item is to "test". Test your theories as much as you can, and when you find something that reacts in a way you didn't expect, think about what it is doing, watch it (read about it) and work out a way you might be able to test it. After all, who cares what "the top nacho in SEO" says when you have been able to prove for your sites that it's irrelevant?

In my opinion, this is what makes SEO so exciting . Like other things in marketing (and in my opinion SEO can be seen as a kind of marketing), there are no fix rules. Watch, think, test, think and monitor: and above all - you need to have a framework that lets you do these things systematically. Looking back at my marketing studies, the majority of the time was spent on perfecting mechanisms to "watch", "test" and "monitor results" -- time spent on choosing the right color, headline or font was minimal (and the choices are irrelevant for a larger picture).

"Monitoring" includes finding relevant metrics to track. Pagerank (PR) is probably not one of those.

Similarly, looking at other sites and trying to extract the influence of individual elements is almost impossible. Let's say you want to create a billboard for a company - would you take technical elements from a Coca-Cola campaign just because they're successful financially? Likewise, creating a billion pages with content won't make you the next Wikipedia. But maybe ... some inspiration from both sides can lead you to an idea that you could test, think about, try and monitor ...

Let's turn it around: if you were a search engine, which technical elements would you value and trust at the moment (knowing that there are professionals who want to game the system)?

John

[Black_Knight]

1. A good offering (products that sell themselves need less of a push, and build more natural links, WOM, etc).
2. Lean, well coded pages with semantic markup and high accessibility.
3. Excellently written copy for the market.
4. Intelligent site structure, reflected in usable URLs.
5. Dedicated ongoing SEO tweaking and promotion.

[JohnMu]

Nice list, Ammon. Good, "common sense" items that we should always think about.

I wonder what you all think: Let's just look at on-page factors; when comparing a site that was "just" well-written, uses intelligent semantic markup, is accessible and usable to one which was technically tweaked and optimized especially for SEO, how much advantage would the tweaked one have, in the short term or long run?

I can't fight the feeling that 99% of the sites that want SEO just need to get things technically and intelligently in order on the pages themselves (semantic markup / accessible / usable) and really don't need any of the extensive tweaking. Additionally, I am always worried that too much tweaking could end up being a liability for a site in the long run.

Does it really make sense to tweak the maximum out of on-page SEO, or would clean content be a better investment for a small/medium sized company that doesn't have the budget for constant tweaking/monitoring?

John

[Black_Knight]

I wonder what you all think: Let's just look at on-page factors; when comparing a site that was "just" well-written, uses intelligent semantic markup, is accessible and usable to one which was technically tweaked and optimized especially for SEO, how much advantage would the tweaked one have, in the short term or long run?

It depends a lot on the market, John. There are a few SERPs where almost all of the top sites are pretty well coded, and well marketed. At which point, the SEO tweaks become all important.

There are a lot more SERPs though where the sites are all universally quite poorly structured, graphically based, with poor use of textual copy, and get along simply on the strength of the brand and backlinks, making it an easy place for the affiliates to come in and take top slot.

There are very few sites hitting 5 out of 5 from my list. Those few do very well indeed.

[Jexley]

Most important things to take into consideration:

1) Is the site worth a toss to a human? (Content)
2) Is the site worth a toss to a bot? (Code)
5) Who do you want to like the site? (Target Demo)
3) Who does like it? (Backlinks)
4) How strongly do they like it? (Quality of links)

I could probably shell out all that I think about this over a few pages, but somebody said something once like "Experiment. Analyse. Adjust."

I forget who (Sullivan? Scuba?) but I like it and have actively lived my SEO life by it even before I heard of it. Sounds like the other fellas agree, just said different.

[Tiffany Hua]

Would you please show us the urls that without metas and PR5+?

[Black_Knight]

Would you please show us the urls that without metas and PR5+?

LOL

You are on one right now. Go to the forum index page and view source:

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 
<html xml:lang="en" lang="en" xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="content-type" content="text/html; charset=iso-8859-1" />
<link rel="shortcut icon" href="favicon.ico" />
<title>Cre8asite Forums (Powered by Invision Power Board)</title>
 
<style type="text/css" media="all">@import url(http://www.cre8asiteforums.com/forums/style_images/css_35.css);</style> 
 
<script language='javascript' src='/spelling/spell_invision.js'></script>
</head>

Its TBPR5, and has no Meta description or Meta keywords.

That was easy wasn't it?

[A.N.Onym]

Since PR depends on the amount of links, meta tags are clearly not important in this case. I see why anyone should prove the point with examples (but I can point you to my blog homepage, too )

Not that meta tags are not useful at all (they are, but not for search engine positioning and not by a large margin).

[niemi]

Would you please show us the urls that without metas and PR5+?

Of course, this is the worst that comes to mind: http://hjem.get2net.dk/j-larsen/

Thanks for the great answers everyone.

I know that SEO starts with the quality of your product, but what I would like to know is how do Google and company measure the quality of a product and thereby its ranking. Technically. Not diluted philosophy. What do they take into account first and then next?

So lets talk about links, tags, site structure and so forth, how much is it all worth?

[Black_Knight]

how do Google and company measure the quality of a product and thereby its ranking. Technically.

They measure the number and quality of citations (links).

A campaign is like a big parade.

If your product, the cause of the parade is not actually good, then the only people in the parade are those you paid to be there. The search engines (and indeed potential customers) are like the people the parade passes. Thy'll look up to see what is happening, but won't actively watch the parade unless it looks like a lot of fun and very impressive. But your parade is only as big as you pay for, and you need to pay a lot of people to make it look impressive enough to hide the fact that the product is just average, or worse. Every inch of progress has to be paid for, and the moment you stop paying, the parade stops happening and everyone stops watching.

With a great product, out of any group of people just seeing it, a few genuinely want to hold a parade to celebrate it (people would link to it, give it good reviews, and talk about it for free). So you pay for giving the parade a good start, but once that parade is going, the people the parade passes look up, and some get involved. The parade grows and grows on its own merit. The money you spend on this parade is purely going into cumulative growth, giving it a head start, and the parade keeps going at a stable pace even without any spend. Your investment in promotion is purely letting it leap ahead, giving it an extra surge.


The search engines don't just look at the number of links. Link weighting is used, so that the quality of the links (citations) is far more important (PageRank, etc). In some cases, the topical expertise of the citation is also important (link from a site that already has relevance for similar terms (Hubs and Authorities). Indeed, search engines are now looking at the Trust value of a link, in terms of downgrading citations from people who are known (or strongly suspected) to have provided good citations for stuff for a bung of cash in the past, or for their mates, not just on merit. They give the highest weight to citations from unimpeachable sources. In other words, the most valuable links are those that are impossible to buy, beg or steal. The best links can only be gained for excellence of the product.

That doesn't mean people can't throw enough effort into a sub-standard or average product to make it seem better than it is. But it takes more effort than with a good product.

This is business. If by making a $1,000 change to a product, you can make your $150,000 a year promotion budget achieve double the ROI, only an idiot won't do it. At least, so we'd like to think. But sadly in the real world, there are a lot of idiots.

A lot of businesses get into a mindset that the only way to achieve double the sales is to show it to twice as many people. They don't think about the conversion rate (how many people shown the thing say no, and how to make a higher percentage say yes - probably that $1,000 change I mentioned). They don't think about the power of Word-Of-Mouth and Viral Mrketing because they don't and won't trust what they can't control 100%. They think that any review they need can be bought, or ignored.

But the role of an outsource in promotions is to be smarter than them. That's your job. They don't hire people who can only tell them what they already know. Your job is to know what will have the biggest effect in marketing and promoting the product. Your job is to know that the product needs a $1,000 change made, and that throwing an extra $150,000 into promotions won't be quite as effective as making that $1,000 change to the product, but you'll play it however the employer decides. Its his money to spend.

[ukdaz]

A campaign is like a big parade.

Quite simply the best analogy I have seen all year.

Me, I tend to use the analogy of growing a vegetable garden when it comes to SEO with clients. Parades seem more exciting though!

Daz

[doogers]

backlinks and new content are key

[Halfdeck]

Top 5 aspects of SEO? Here's my list:

1. Think of an original idea. Do this before anything else, before you touch your meta tags, before you decide on a cool domain name. Building the 50th YouTube is a waste of time.

2. Your website isn't a surfer-trap - its a product. Would Microsoft release HD trailers for Halo 3 if the game wasn't past the design phase? Don't try to build the trailer first and the game later.

3. Optimize for Google. Done building a great site users will love? Then you gotta wire it up for the search engines.

4. Spy on your competitors. Your goal is to build the best site there is. How do you win American Idol? Be the best singer/entertainer/performer. Same thing with Google. To be the best you can't work in a vacuum. Keep your eyes on your competitors because SEO is a race without a finish line.

5. If you just build it, they won't necessarily come. (but if they come, you better have something that'll knock their socks off.)

Edited by Halfdeck, 23 August 2007 - 11:53 PM.

[ChuckTM]

Top 2

1. url
2. Relevant, unique copy about one subject. 6 pages long

That's it.
If you do that, you will get links.
(Even from your competitors)

Ratio?

keyworded URL = 20%
keywordedPage = 40%
keywordedLinks = 40%

You can make due without, but all three add up to the best results.