How can you not see the flaw in what you said. You claimed that the one site was ranking only because it also had a link pointed to the page with the phrase in proximity to the link, and that's what made that page different. All three phrases have that. It should be readily apparent to someone not suffering from wet brain that the only phrase that ranked is the one with the phrase in the meta description. -Michael
Mvandemar is giving some good points. This is getting interesting. 7 pages, aye? I like starting long topics.
KEY TOOLS? No way is meta keywords a "key tool" of Google. I doubt it is used at all, but saying it is a "key tool" implies it is more, as, or almost as important as content or links.
Almost a month later, and the test is still holding up. None of the other terms ever ranked. The only one that did is the one that was included in the meta description. -Michael
Google does uses Meta description. But then it makes sense for them to use it. Using a good meta description increases the chances of getting that click from the search engines. For the rest I'm not sure.
I still don't believe that meta description has any influence over search results, I just think it's the description.
You're right. I previously pointed out the flaws in mvandermar's little "test". He has difficulty accepting that his little "experiment" just didn't "prove" what he wanted it to show. I've also previously advised him to do a little reading into experimental design and sampling but apparently he's too busy defending his little "tests" to follow through on that. Expect him now to pop in and try to insinuate that one or both of us has a drinking or drug problem. That's his typical response to people who don't accept his statements at face value.
Re how to tell Google to not use the ODP description when their description no longer matches the content of your website, use this meta tag: <meta name="googlebot" content="noodp">
No, you didn't. When asked why the only one of the three phrases gained rankings, you replied it was because it was the only one that also had the phrase in proximity of a link to the page. All three phrases were used as anchor text, and all three phrases were in close proximity to all three links. Additionally, there is now another site with one of the other phrases "in proximity" to a link pointing at the same page. Doesn't matter. To date the only phrase that is ranking is the one used in both the anchor text and the meta description combined. You also said that it was just because it hadn't been "long enough" yet: What about now? Long enough for you? -Michael
Sorry, no. It still doesn't prove what you claim it proves. Follow my suggestion to read up on experimental design - you might learn something. The only clear way to demonstrate that Google uses the meta description tag for indexing and ranking would be to show me a page that ranks for a term that ONLY appears in the meta description tag. And as I said elsewhere, tests using that approach in the past have been negative. Learning a little about experimental design will help you to avoid the problem of confounding variables and intervening variables.
As I've already stated, that's not what is being proved here. It takes X amount of a certain factor in order to appear in a query for a given phrase. We already know that having the phrase in either on-page text, in the title, or in the url automatically meets that threshold, but just having it in meta description or in the anchor text is not enough. However, as I have shown, when you have it in both the anchor text and the description, it suddenly meets that threshold. It bypasses the new anti-link bombing measures that they put into place. It actually is one of the factors that they use to determine whether or not to show the page. I'm not saying that it's a huge factor that will make you outshine all of your competitors, and my guess is that if having the phrase in the title gives you a ranking factor of say, oh, 3, and having on-page a couple of times gave you a factor of 1.2, then having it in the description is probably somewhere in the .01 range... I was just saying that it exists. -Michael
If it does, you have failed to prove that. One more time: Do some reading in experimental design and confounding variables. That's not what you demonstrated at all. All you've managed to show us is what we already knew - that anchor text works. If you don't isolate the variable/factor, you've proven nothing.
One last time... if it's just because of the anchor text, then why is only one phrase ranking? -Michael
*sigh* I don't think you get it and I don't think you want to get it, vandermar. Go do the reading I suggested. That may help, although I'm beginning to doubt it.
SEO aside.... Google does use it and it shows up in the SERPs. People searching the internet do read them before clicking on the site. The more attractive the meta description the more clicks. To solve this test theory why not just change one of your homepages meta description and add a phrase that returns 0 results on Google. Wait a few weeks and if it shows up then you know it does count it for SEO purposes.