Test, Test, Test

I'm going to borrow this quote from Seth Godin, who borrowed it from Kevin, who borrowed it from The Count Of Monte Cristo:

"I have been told," said the count, "that you do not always yourselves understand the signals you repeat."

"That is true, sir, and that is what I like best," said the man, smiling.

"Why do you like that best?"

"Because then I have no responsibility. I am a machine then, and nothing else, and so long as I work, nothing more is required of me."

In SEO, what do people say works vs what actually works?

Filthy Linking Rich

If you haven't already seen it, check out Mike Grehan's Filthy Linking Rich from 2004. It's as relevant today as when it was written. Mike makes the point those who are already rich, tend to get richer. Those sites that have the most links, tend to get more, because those sites have the wealth of exposure already.

This is why it can be tough to get a new site ranked.

Those sites that are link poor, no matter how great they are, will struggle to be found in the search engines. "If you're great, people will link to you" is not necessarily true because a link-poor site is unlikely to show up in the search results in the first place. Initial discovery will likely happen via other means.

Search Engines Don't Care About "Great"

The search engines don't reward information that is great. The search engines reward information that is popular, or appears on a site that is deemed popular.

If your aim is high rankings, then it could be argued it is better to focus on being popular, than it is to focus on creating quality. Look at a lot of the content on mainstream media news sites. Is such content really of higher quality than other sources, or does it just happen to appear on the right domain? If such content wasn't published on a popular domain, and was published on a brand new site instead, would it ever see the light of day?

When it comes to search engines, it really does matter who, not what, you know.

Test, Test, Test

People often repeat what they've heard.

I'd urge you to test, if only to be aware of the level of misinformation you're may be getting from SEO forums and blogs. There is a lot of "thuthiness" bouncing around the SEO echo chamber. But how much of it is based on evidence?

Challenge SEO punditry. By testing.

Search on a keyword phrase. If you search on a high volume phrase, chances are you'll see a page ranked at the top based largely on the link profile of the *site* on which it appears. The site will have many links, and this link value filters down through the pages. A few positions down the SERP, you'll likely see pages based on their inbound links, even if the site on which they appear doesn't have many links.

Take a look at the back-links.

How many of the sites you're seeing have backlinks that are clearly autogen? Blog spam, forum spam, etc?

Chances are, you might find quite a few.

I'm looking at a product-oriented serp right now that has Wikipedia at the top, followed by the brand holder of this product, followed by a site that has tens of thousands of auto-gen inbound links in position three. I kept scanning through the links until I found what I considered to be a great match to my query.

On page five.

Now, what I judge to be good might not be objectively great, of course. I've made a subjective judgement, just as Google has made a subjective judgement. Try it out yourself. Rather than rephrasing a query, scan through the pages until you find a page that does answer your query.

Then evaluate the sites above it. What, exactly, are they doing? How many of them are doing anything more complicated than "getting a lot of links"?

For all the fluff about 100's of ranking signals, it still appears that mass link bombing, from rubbish sites, works a treat.

Simple Testing

Am I talking nonsense?

One way to find out.

For those new to SEO, be wary of what you read. A lot of it is conjecture. What the old skool SEO's used to do, and the more serious SEO's still do today, is test for themselves, as opposed to relying on the pundits.

Testing can be done with existing tools, like the SEO Toolbar - and the tool set for members. Little plug there ;) There are a huge number of tools around, but one of the most important is a tool that will allow you to analyze link structures.

Grab one of these tools and go through the sites you're competing with, and pay close attention to the backlink profile of both the root domain and the page that is ranking well. Make a note of what is working, without making a moral judgement about the validity of the techniques being used.

You can also test with throwaway domains. Register a new domain, for an obscure keyword within your niche, and try and isolate the effects. Point one link at the domain, see what happens. Point ten links at it. What happens? Point links from a variety of domains. What happens? Change the link text. What happens?

Simple stuff, right. But simple stuff that will teach you much more about SEO than reading the pundits blogs and tweets today.

Including mine :)