Meta Keywords still do matter

[mike2010]

On one of our sites, for 6 straight months I couldn't get us to appear anywhere on google for a search of our domain name. So I put it in meta keywords (just the domain name , and without the extension) a month ago...and within a week we went to #1 for a search on google of our domain.

I was recently baffled by this clip from Cutts saying they don't use meta keywords anymore. Then again, it is from 3 years ago -

[youtube.com...]

Any confirmation from Google recently that they are once again starting to put emphasis back into meta keywords ?

I could tell you first hand from recent experience, that they do still matter.

[varun21]

hmm, there could be some pattern to it. Maybe they ignore all meta keywords but in case there is only one meta keyword then they'd give it weight?

[JohnNZ]

Interesting result. Although, personally, I've given up trying to work out 'what works' with Google these days, too many things going on in the algo.

I guess one way to confirm it, is to remove the meta keyword and see if your site disappears :-)

[Sgt_Kickaxe]

Google doesn't trust any signal that is provided by a webmaster that easily, especially keywords which I doubt play a role with Google anymore. When your site gets reviewed by a human(or 12) and receives good grades then they'll begin to trust your site.

[BeeDeeDubbleU]

Looks like coincidence.

[FranticFish]

1) Put a unique, gibberish word (that there are no results for in Google) into a keywords metatag on an indexed page.
2) Wait until it's reindexed.
3) Search for that word in Google.
4) Post back if Google shows your site as a result.

[indyank]

That isn't a good test either. I know that most SEOs do those kind of tests.

But the fact is, Google tries to show its users at-least one relevant result for their search queries. If that gibberish word is found in the source code of any html page, irrespective of whether it is defined within a meta keyword or not, Google might still return that page as the most relevant result for the query.It however doesn't mean that meta keyword tag is always used by them.

[indyank]

However the OP's observation is really interesting though the chances of it being coincidental is high.

[indyank]

I guess one way to confirm it, is to remove the meta keyword and see if your site disappears :-)

I think that is a much better test. :)

Unfortunately, Google now seem to have a rank modifying patent to deal with that as well.

[FranticFish]

That isn't a good test either

Wrong.

If that gibberish word is found in the source code of any html page

Er, then it's not a proper test! That's why I said...

Put a unique, gibberish word (that there are *no* results for in Google)

If the only source of that 'keyword' on the entire web is one metatag in one webpage and Google returns that webpage as a result, that'll do for me.

[indyank]

a unique gibberish word is exactly why I feel it isn't a good test for me....

[aakk9999]

Putting a gibberish word which currently has no results in Google into meta keywords element on the page and then seeing whether the page ranks is a different test to what has happened to mike2010.

Mike2010 claimed that meta keyword helped his page rank where the keyword was his domain name, therefore it was "mentioned" elsewhere on the site.

Better test would be to pick the page and a keyword for which you are not ranking well, then add this keyword to meta keywords and see if the ranking for that keyword improves.

[edited by: aakk9999 at 4:19 pm (utc) on Sep 1, 2012]

[BeeDeeDubbleU]

It may not be a good test for you but it would be proof that they read meta content if that is the only place it appears.

[aakk9999]

I am sure they read it, it is just whether they use it as a signal and how.

[indyank]

I do agree that the OP's observation is interesting but he has to add more. Is it a keyword based domain name or a unique brand name like say, "Google"? I am almost sure that it is the later and if so, the site ranking for its own unique domain name shouldn't be a surprise as it is live for 6 months+. The OP should be more worried about why it wasn't ranking for that unique brand name so far and the meta keywords thingy looks more coincidental to me.

[mike2010]

Mike2010 claimed that meta keyword helped his page rank where the keyword was his domain name, therefore it was "mentioned" elsewhere on the site

#1 it wasn't a popular phrase at all , thus the reason for it to be pulled from nowhere to be found (in google).... to #1.

#2 Unless you have a really good 1 word domain name, I don't see a reason why people would place their domain name in other parts of their site when it's 2 or more words...and placing it without a space. That domain was 2 words. So I did not have it mentioned anywhere else on the site. Not in meta description, not in title, not in the text on the site or anywhere. Placing it in meta keyword was the only place it was placed. And like I said, within a week it boomed to #1.

Just that single instance did it. You'd figure since it was an unpopular phrase and part of the domain, google would automatically give it weight for rank. But it did not. Until placed in meta keyword.

I'd recommend others try the same if they have a domain hangin around for testing.

thinking about removing it and seeing if rank disappears again..

[indyank]

I find it interesting that you chose not to mention your domain name in the title of your home page as well!

If i understand what you say correctly, it is an unique phrase that might not be found elsewhere on the web. Is it true?

Is the rank that you see now for the 2 words together without a space or with a space between the two words?

So the domain url was the only place where google found it and that IMO isn't enough for google to automatically give it some weight and rank your site. All it needed was a mention of that unpopular phrase somewhere in the html source code of the page.

[chalkywhite]

Well, I for one am going to test this, I rank 7th for a keyword and have done for a year or so. I do not run keywords at all. Just entered the keyword now. Lets see.

[Robert Charlton]

meta keywords still do matter

Only in an oddball case like this one where there are presumably legitimate links pointing at the page, but the text isn't reflected elsewhere on the site. The situation illustrates how Google's anti-Google Bomb filter works, and it tells us that an occurrence of the term in the meta keywords will disable the filter.

I discussed a similar example a couple of months ago, in this July 2012 thread....

Meta keywords - are they worth it?
http://www.webmasterworld.com/google/4471188.htm [webmasterworld.com]

As I'd noted about an example we allowed to be posted...

The page is ranking because of inbound links along with years of quirky online marketing. Several of the meta keywords appear to be supplying a confirmation to Google that the inbound links containing them aren't unwelcome Google bombs, but the meta keywords themselves are not otherwise causing the page to rank.

Some years back, I remember that on a search engine conference panel, when asked about meta keywords, Yahoo said that it referred to meta keywords only when all other signals had been exhausted. On the same panel, Matt Cutts said that Google wasn't using meta keywords, but that it reserved the right to do so when it wanted to. ;) In the case of disabling the Google Bomb filter, this might be such an instance....

If you remove the meta tag without adding confirming text on the page, chances are that the site will disappear again. This doesn't mean that meta keywords help in a competitive ranking situation... just that when you deny Google all rational signals of identification, they will use meta keywords as a confirmation. No search engine is psychic. It's really best if you help them by including your relevant content on the page.

Worth noting in this regard that way back, when meta keywords were given weight by search engines, Inktomi considered it spam to include words in the meta keywords that weren't also included onpage.

Regarding this test...

Put a unique, gibberish word (that there are no results for in Google) into a keywords metatag on an indexed page.

I'm curious whether a meta keyword by itself will be enough. It might need the boost of an inbound link (in which case there would be two pages ranking for it), but I'm not sure.

It would be an interesting two-step test.... do it first just as the meta keyword, then, if there are no results, hit it with inbound link(s). Also, I'd monitor also with a site:testdomain search at various stages of the test, to see whether there are thresholds of ranking that a meta keyword can trip.

Well, I for one am going to test this, I rank 7th for a keyword and have done for a year or so. I do not run keywords at all. Just entered the keyword now. Lets see.

IMO, it's not likely that the meta keyword will help in a competitive situation at all.

[FranticFish]

If we're quoting that thread please can it be noted that the phrase mentioned WAS actually found in a number IBLS to the page being discussed! See last post.

[Robert Charlton]

FranticFish - Yes, to be clear about it, I'd also found the phrase in several IBLs... and I feel that the primary ranking mechanism was via inbound links (essentially a Google Bomb).

The meta keyword simply confirmed to Google that it was "OK" to turn off the filter that protects against mischievous inbounds.

[FranticFish]

It would be an interesting two-step test... do it first just as the meta keyword, then, if there are no results, hit it with inbound link(s)

Robert - one IBL by itself might be enough for a page to rank, even for a gibberish phrase - at least this was the case when I tested a few years back. So I'd suggest a 'control' page that is linked to from the outset with the same phrase.

If that is found from the start then it makes the job of proving that it's the COMBINATION of keywords tag and IBL that's producing results far harder to do, if just a link on its own will suffice.

[mike2010]

they will use meta keywords as a confirmation.

Good way to put it. If they expect that keyword to be elsewhere on the site, but isn't... they'll use 'meta keyword' to confirm and give it weight. Especially since it was part of the domain and not a popular phrase.

I find it interesting that you chose not to mention your domain name in the title of your home page as well!

I didn't solely because it's not a popular term, and obviously without the space, very few would search for it like that. But I do have it on other parts of the site WITH the space.

example : if my domain was dogcat.com , I put dogcat as the meta keyword. (without the space)

[RegDCP]

I have always wondered if Matt Cutts' comment was accurate when he said that they don't use it for search ranking.

Properly used metas are normally a part of database markup.
They are "Information about information.", which means they accurately describe the content of the page.
This should be as concise as possible.

The text strings in Meta tags are weighted.
The fewer the words, the more weight each carries.
The closer to the beginning of the string, the more weight.

When MC told us that the KW tag was not used, I doubted that Google would throw away such a valuable part of an indexing process.

I suspected that this tag was still of great use in determining over-optimization.
If it was concise and accurate it passed the test.
If it was stuffed with every keyword imaginable related to the content, the page got a slap down.

Are the results repeatable?
Can you take another low ranking page, add a kw tag, and get an improvement?

[Robert Charlton]

RegDCP - Google probably doesn't throw away anything, but certain signals aren't worth the computational overhead or the noise they might create.

Your points about "properly used meta tags" for information retrieval are good ones. In competitive search, though, meta keyword tags have been used ineptly and improperly for so long that Google probably decided at a certain point that they weren't worth the overhead.

Whether they're used for algorithmic spam detection is debatable, but I'd guess not, as so many people think that meta tags are pie in the sky wish lists that Google would probably be getting noisy spam signals.

That said, in public site review sessions, I have seen Matt himself use meta keywords tags (and comment tags) as indicators of user intent. He almost always looks at them for evidence of keyword stuffing (particularly for the four p's), which suggests that they are looked at in Google spam team manual reviews as well.

[RegDCP]

If Matt looks at them for instances of KW stuffing, wouldn't Google's agos?

Easy enough to program, (if keywords exist on page) = Then no action.
(If keywords not exist on page) = debit xx positions.

[Robert Charlton]

If Matt looks at them for instances of KW stuffing, wouldn't Google's algos?

Matt looks at these in public presentations and chuckles along with the audience. There is some entertainment value in this, which also serves to bring a point home.

Again, read what I said about noisy signals.

Easy enough to program, (if keywords exist on page) = Then no action.
(If keywords not exist on page) = debit xx positions.

And zero collateral damage, right?

[jimbeetle]

(if keywords exist on page) = Then no action.
(If keywords not exist on page) = debit xx positions

Sounds like what Alta Vista was said to do so many moons ago.

As for what Yahoo did:

So, if you have a product which is frequently misspelled. If you're located in one community, but do business in several surrounding communities, having the names for those communities or those alternate spellings in your meta keywords tag means that your page is now a candidate to show up in that search. That doesn't say that it'll rank, but at least it's considered. Whereas, if those words never appear then it can't be considered.

Mike Grehan interview [e-marketing-news.co.uk] with Jon Glick back in 2004. (About 2/3 of the way down the page.)

[skibum]

I don't think they carry much weight but don't think Google ingores them either.