SEO Experiments

[Randy]

A big part of becoming good at SEO is experimentation. You shouldn't do anything SEO-related just because someone tells you it's a good idea, and you should be skeptical about any advice you receive from anyone on a forum. (Even from me. Especially from me, in fact. )

But a lot of folks don't know how or don't go to the trouble of conducting any SEO experiments. Here are some suggestions I would make to someone who wants to conduct an SEO experiment but doesn't know how:

1. Start with a hypothesis. A hypothesis is just a guess at a cause and effect relationship related to your SEO. For example, your hypothesis might be: Including my target keyword phrase in the body copy on the page will make my page rank better in Yahoo.
2. Pick 5 or 10 pages on one of your sites to experiment with. Then change only the one single thing you want to test. In the previous example, you could remove every single instance of your target keyword phrase from the body copy of those 5 to 10 pages.
3. Observe what happens. It might take as much as 2 weeks for the the pages to get recrawled and reranked. In the previous example, if the pages all seem to move down in the rankings after you removed the target keyword phrase from the body, then chances are your hypothesis was correct.
4. Change it back to the way it was before. Then watch the results again. If you remove the target keyword phrase from the page, and your rankings go down, then you re-add the target keyword phrase, and your rankings go back up, then you can be pretty confident that your hypothesis is correct.

Based on how significant the changes are in your rankings based on the isolated changes you make when you're experimenting, you can get a sense for how much weight various factors have on your ranking. For example, if you drop an average of 10 spots in the SERP's by removing the target keyword phrase, then move back up 10 spots, you have a reasonable idea of how important that factor is. If you test another factor, and it results in an average 20 spot move up or won, you'll know that particular factor has about twice as much weight in the search engine algorithm.

[mjmurphy53711]

I do this a lot, I'm always changing things to see what happens since I have no seo knowledge to speak of, I'm learning by doing.

[Randy]

Originally Posted by mjmurphy53711

I do this a lot, I'm always changing things to see what happens since I have no seo knowledge to speak of, I'm learning by doing.

You'll be better at SEO than most by learning that way.

[mypokercorner]

Originally Posted by Randy

You shouldn't do anything SEO-related just because someone tells you it's a good idea, and you should be skeptical about any advice you receive from anyone on a forum. (Even from me. Especially from me, in fact. )

I stopped reading your post from this point on, but I would be very interested to hear your toughts on what not to do SEO wise.