Forum Moderators: open
So what i did was: Took my most important terms and put those inbetween commas. Then followed that with the not so important words and separated them with the traditional space. My site was on the third page and now I cannot see it on the first 20 pages!
What do you think? Could this be harming my site in a huge way or do you think google is merely processing/reviewing the changes and i should maybe give it more than the two days waiting that i have done so far(Impatient am i).
Look forward to your valuable replies:)
Happy Surfing
60 days ago I had a site showing the meta content for the G SERP 'title' and SERP 'description'. Cached page proved only location of the words/phrases to be in meta tag(s). Though it does not matter, the words/phrases showing in the results were never located anywhere else on the page.
Google does not read meta keywords.60 days ago I had a site showing the meta content for the G SERP 'title' and SERP 'description'. Cached page proved only location of the words/phrases to be in meta tag(s). Though it does not matter, the words/phrases showing in the results were never located anywhere else on the page.
It does read meta descriptions.
Any indication of a weight on the meta description? Anyone?
Meta Description is still a must for crappy Engines and appearance. Meta keywords are useless.
I need to clarify this for the others. The Meta Keywords is not useless. Yahoo reads the Meta Keywords tag.
Why does Yahoo! read the meta keywords tag...?
Same reason they use any other factor in their algo - they find it useful.
....and how do we know this?
Well, I know it accidentally on purpose. I accidentally left a nonsense placeholder (unrelated to anything on page) in a keyword tag on a couple of pages not so long ago. Couldn't be bothered fixing it when I found it, so left it and then forgot about it. A while later I decided to search it out of curiosity - et voila, ranking in Yahoo. No sign in Google though.
okay, I found it with Google (as usual). Search for <tim yahoo meta keywords>
Any indication of a weight on the meta description? Anyone?
Zero for G. Tested and confirmed with test pages using nonsense word.
I think you should always put in meta keywords for a number of reasons, even for google...
1) Simple spiders being used to make small niche directories may use them. Getting onto these directories, where someone is taking the trouble to spider the web and not just taking a feed from other established directories/engines may be worth while. This niche directory could even give you pr.
2) Testing to see whether google uses meta keywords is IMPOSSIBLE. You have to have the words in the meta keyword also on the page, otherwise at best they will be ignored. Its a simple rule to prevent keyword spamming. The meta keyword is a guide for a spider which has to be backed up by other on page factors.
3) Percieved wisdom dictates that google ignores meta keywords. I reckon they use every scrap of information they can. Its hard enough making a valued judgement with the known data they use, I doubt they totally ignore anything. Think about it, they are in the business of allocating keywords.... do you really expect them to ignore the precise metatag that should provide this? They know it can be abused, just like any other element of a website, but they try to find ways of detecting or preventing any abuse. I suppose a great way is to spread the rumour it is ignored.... which will mostly be heard by seo webmasters!
4) Using spaces rather than commas is probably better because it generates more variation of keyword phrases. e.g. "red, widget, blue elephant," does not show the phrase "widget blue elephant" or "widget blue" whereas without the commas it does.
Do meta keywords have a negative effect with google? I don't think having them could be negative unless you keyword stuff with words not on the page, or you just keyword stuff beyond a sensible amount.
Is it good to have them?
Only google knows, but doing them properly is prudent. Indirectly it is good, because others do use them.
Skylo wrote "Took my most important terms and put those inbetween commas. Then followed that with the not so important words and separated them with the traditional space."
I suspect that google does read them, don't follow the crowd, you may not stand out because of it :) . The problem is not with commas or spaces, but the amount of keywords you have presented. You may have diluted your major keywords with too many minor ones. Having said that, identifying this as the reason for your drop in rankings is never so simple. If you have dropped, check your backlinks and the pr they could be supplying first. Reverse the changes you made and wait a month to see what happens. Or, better still, reduce the keywords in the meta keyword tag to just the very important ones, that are reflected as such on your page.
Zero for G. Tested and confirmed with test pages using nonsense word.
It is a difficult one to test. Of course, keywords in the meta keyword list could be given more weight in the body when performing their algorithm. So if you have the word 'web' in your webpage, google could give it a normal 100% weighting, for normal size text. If 'web' is in the page as well, google could well give it a 105% weighting.
The reason I beleive the above to be the most accurate is because this would stop people spamming the keyword meta data. The words would also have to be on the page itself.
Unless you tested the nonsense word with the word in the page body and also in the meta keyword tag, then the test is not yet conclusive.
Me, am doing just fine on G and Y! without them. Risk big G rankings for a theoretical possibility that few crappy bots/directories may still use this ancient method of determining what the site is about? Well...
2) Testing to see whether google uses meta keywords is IMPOSSIBLE. You have to have the words in the meta keyword also on the page, otherwise at best they will be ignored.
Not impossible, quite easy in fact. Just as you state, include it in the meta tag and in the body to see if the combination results in a higher ranking than either factor individually
Unless you tested the nonsense word with the word in the page body and also in the meta keyword tag, then the test is not yet conclusive.
Having the keyword only in the meta tag and not on the page results in that page not appearing in the SERP for that keyword, therefore I conclude that the value is zero.
Do I include it in the meta on my sites? Sure, for the benefit of other SE that might still use it and the remote possiblity that having it along with it in the body in combination may have some value, which I have not bothered to test for.
"... include it in the meta tag and in the body to see if the combination results in a higher ranking than either factor individually"
The trouble is that during this test other things can change. A new link may kick in or you hit a different datacenter that has just applied a new algo or even other sites may rise or fall. Unless google stays static, you cannot be sure your test is conclusive. I believe google mixes the results with different algo's and serps on different data centers to confuse any webmaster's tests. We know anchor text and pr application can take days or weeks to have an effect, you never know if and when so testing is.... impossible.
The trouble is that during this test other things can change. A new link may kick in or you hit a different datacenter that has just applied a new algo or even other sites may rise or fall.
No new links can kick in because the test is done on a special page in which only I control the in and outbound links. I also control all other pages that rank for this keyword, since it is a nonsense word that I made up.
Unless google stays static, you cannot be sure your test is conclusive. I believe google mixes the results with different algo's and serps on different data centers to confuse any webmaster's tests. We know anchor text and pr application can take days or weeks to have an effect, you never know if and when so testing is.... impossible.
Of course Google is always tweeking the algo. That doesn't mean testing is impossible or non-conclusive. The results can be conclusive for that point in time and in fact the results can and have remained the same, and will continue to do so until the next tweek.
Naturally we can never say that what is important now, will continue to be so forever in the algo for Google, but we can control the variables and design tests that conclusively show the relative importance of various factors for SEPR at this point in time.
Having the keyword only in the meta tag and not on the page results in that page not appearing in the SERP for that keyword, therefore I conclude that the value is zero.
Interesting. But that proves only one thing - Google ignores meta keywords if they do not ALSO appear on the page (i.e. it assumes that keyword in the meta tag is spam or no longer applicable). But if it appears on the page as well, then it is difficult to test - it may be minor, but it could weight words in the body slightly higher if they are also listed in the meta keywords tag. To perform the test properly, if can ever be done properly, we need an identical page with the meta tag and without the meta tag - both must have the keyword in the page body (as would likely be the case with genuine meta keywords).
Google is constantly changing the algorithms and testing is very difficult and usually 100% proof never exists on anything!
Of course, it is highly unlikely that meta tags will cause any lowering of rankings in Google. It also seems that it does not improve rank. But that is not the case in other engines, even Inktomi. So putting in these tags does seem worth the few minutes it takes to do so so.
PS. If I were writing a search engine algorithm, I would ignore meta keyword tags that did not appear in the page because of spamming. Wouldn't that make sense?
But that is not the case in other engines, even Inktomi. So putting in these tags does seem worth the few minutes it takes to do so. I would ignore meta keyword tags that did not appear in the page because of spamming. Wouldn't that make sense?
Absolutely. That pretty much sums up the use of the META Keywords Tag.
It's still impossible. You need to have 3 identical pages running at the same time with the three variations of the nonsense word, e.g. one with the word on the page, one with the word in the metatag and one with both. If the pages are identical you will have duplicate content issues, so you will need to know how google will discriminate. If the pages are not identical, then there could be page size issues, position or word density. Linking the pages 'identically' is also very difficult, if all the links came from the same page then what text do you use? there may be a relationship between the keyword as a link in + in the metatag+in the body. You can't have 3 identical anchor text from the same page as that could cause problems and using different text or image links may cause you missing the relationship between anchor and metatag keyword. Use different pages to link to it and you have to know their exact pr is all identical.... impossible. What about links from the same ip, internal links may be treated differently and ignore metatag keywords, external links in may use them!
The list of variables goes on, and if you introduce time, the effect of fresh content having a short term boost etc. you have a new set of unknowns. Did one datacentre pick it up as fresh content and another didn't?
And then, in the middle of it all on a longer term test, another spider finds the page and a nonsense directory lists you!
There are too many potential variables to be sure. And another issue is using a nonsense word.... there will be no broad match relationships, no authority sites, no hubs etc. etc. The use of a nonsense word makes the test non conclusive in itself, because the serps will have no theme.
I personally am very unsure that google uses keyword meta's but "never say never".... and never assume you even know all the variables. They change daily in subtle ways, even for a nonsense word.
But if it appears on the page as well, then it is difficult to test - it may be minor, but it could weight words in the body slightly higher if they are also listed in the meta keywords tag. To perform the test properly, if can ever be done properly, we need an identical page with the meta tag and without the meta tag - both must have the keyword in the page body (as would likely be the case with genuine meta keywords).
Not difficult to test at all. As you state 2 pages set up identical except with and without the meta tag (plus a third with only the meta tag, and some content shuffling to avoid duplicate content.) I haven't bothered to test this simply because as a matter of course, I always include the meta tag keywords anyway.
It's still impossible. You need to have 3 identical pages running at the same time with the three variations of the nonsense word, e.g. one with the word on the page, one with the word in the metatag and one with both. If the pages are identical you will have duplicate content issues, so you will need to know how google will discriminate. If the pages are not identical, then there could be page size issues, position or word density.
Not impossible. All my other test take into account the mentioned duplicate content issues of word count, density and position.
Linking the pages 'identically' is also very difficult, if all the links came from the same page then what text do you use?
Not difficult at all. All links to the individual test pages come from one page, and the anchor link text is simply 1,2,3, etc.--no relation to the keyword.
And another issue is using a nonsense word.... there will be no broad match relationships, no authority sites, no hubs etc. etc.
Exactly, these other factors are eliminated from contaminating the test results. So results are purely based only on the on-page factors that can be used to compare the relative impact of individual page design elements.
It puzzled me at the time because there was plenty of relevant text on the page for Google to use as a ransom note. Why they were showing the Meta keywords I do not know.That sounds like the page has a bit of hidden text somewhere. Often hidden text is simply keywords and has the appearance of, if not being cut and paste from, the meta keywords. Also check for alt text within an an image link.
So the test is flawed.
1) It does not test for links from other sites. These will be treated differently.
2) It does not test anchor text relationship to the meta tag.
3) It does not test for hilltop.
I'm sure there are more, but the above alone would render your test inconclusive.
