homepage Welcome to WebmasterWorld Guest from 54.166.122.65
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member

Home / Forums Index / Code, Content, and Presentation / HTML
Forum Library, Charter, Moderators: incrediBILL

HTML Forum

This 55 message thread spans 2 pages: 55 ( [1] 2 > >     
Metatag Labyrinth
Which ones are the most important, and which ones are just good to have?
neophyte

10+ Year Member



 
Msg#: 3409150 posted 2:34 am on Jul 31, 2007 (gmt 0)

Hello All -

For the past two years, I've had a standard template of Meta tags which I have dropped into every project. I got this tag list about a year ago by viewing the source of #*$! and just copying/pasting/altering to my needs.

<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
<meta http-equiv="Content-Language" content="en-us" />
<meta name="ROBOTS" content="ALL" />
<meta name="Copyright" content="Copyright line here" />
<meta http-equiv="imagetoolbar" content="no" />
<meta name="MSSmartTagsPreventParsing" content="true" />
<meta name="description" content="Description here" />
<meta name="keywords" content="keywords here" />
<meta name="Rating" content="General" />
<meta name="revisit-after" content="15 Days" />
<meta name="doc-class" content="Completed" />

After reading online about a lot of changes in how important/not-important certain meta tags are now-a-days, I'm now wondering which ones on the above list should stay, which should go, which should be added.

I don't know if this is a good idea or not, but I'd really rather just stay with a stock-list of meta tags that I just drop into each project rather than re-writing this list from scratch for each project.

Any and all insight greatly appreciated!

Neophyte

 

encyclo

WebmasterWorld Senior Member encyclo us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 5:00 pm on Aug 1, 2007 (gmt 0)

Looking at each one in turn:

<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">

The most important one - you must always define the charset on every page, and using the meta element above is the most common way. Place this before the title element please. ;)

<meta http-equiv="Content-Language" content="en-us">

NOT required. You should define the language with a lang attribute on the <html> opening tag: <html [b]lang="en"[/b]>

<meta name="ROBOTS" content="ALL">

"ALL" is the default, so this is NOT required unless you want anything other than to allow indexing. Other values are noindex, noarchive, nofollow.

<meta http-equiv="imagetoolbar" content="no">

Not required, but useful if you have larger images on the page, as it stops the annoying context icon in IE.

<meta name="MSSmartTagsPreventParsing" content="true">

NOT required, a MS technology which was not implemented.

<meta name="description" content="Description here">
<meta name="keywords" content="keywords here">

Description is important - have a unique one for each page not a general one for the whole site. Search engines like it and use it occasionally for the snippet under the search result. Keywords is of little value, but may possibly have a minor effect.

<meta name="Copyright" content="Copyright line here">
<meta name="Rating" content="General">
<meta name="revisit-after" content="15 Days">
<meta name="doc-class" content="Completed">

Code bloat: NOT required. Unless required by an internal search engine, they have no value.

Note: you used the trailing slash [b]/[/b]> - you should only use the trailing slash if you are using XHTML, not for HTML 4.

[edited by: encyclo at 6:35 pm (utc) on Aug. 1, 2007]

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 5:06 pm on Aug 1, 2007 (gmt 0)

Place this before the title element please.

I think that is the first time I've seen someone confirm this practice. I would think that is the very first thing the UA needs to see after the DOCTYPE.

My normal template, in this order...

<meta http-equiv="content-type" content="text/html; charset=iso-8859-1">
<title></title>
<meta name="description" content="">
<meta name="keywords" content="">
<meta name="robots" content="robots-terms (to prevent indexing, caching, etc. Not to "all" or "index, follow")">
<meta name="author" content="">

Author is required to meet WAI validation.

<meta name="robots" content="none">

I use the above regularly with great success. Helps me to control which pages are getting the juice. When you reach a certain number of pages indexed, you want to start managing which pages are getting indexed. The above element is a simple solution for the prevention of indexing.

Google will not show a page that contains the above element. Neither will Yahoo! or MSN. As long as there isn't an entry in your robots.txt file, the page will not appear in the index.

encyclo

WebmasterWorld Senior Member encyclo us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 6:09 pm on Aug 1, 2007 (gmt 0)

Place this before the title element please.

It's logical - the meta charset element defines the character encoding, and so you should always define the encoding before any visible content - this includes the title element. This is not a validation issue, but a question of best practice.

swa66

WebmasterWorld Senior Member swa66 us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 6:38 pm on Aug 1, 2007 (gmt 0)

<meta name="dmoz.id" content="..." />
It is (was?) important for getting some RSS feed aggregators to work. But it might be outdated, dunno.

The original poster seems to be using xhtml, as such the encoding should be defined in the very first line of the xml.
e.g.: <?xml version="1.0" encoding="ISO-8859-1"?>

nomis5

WebmasterWorld Senior Member 5+ Year Member



 
Msg#: 3409150 posted 7:35 pm on Aug 1, 2007 (gmt 0)

So if you don't include the meta charset element what is the assumed default by Google and others?

texasville

WebmasterWorld Senior Member 5+ Year Member



 
Msg#: 3409150 posted 7:56 pm on Aug 1, 2007 (gmt 0)

Two of my favorites:

<meta content="NOODP" name="ROBOTS">

To prevent that dry odp description from appearing in the search index.

and:

<base href="http://www.example.com/">

to establish the base url of the site.

adfree

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3409150 posted 9:49 pm on Aug 1, 2007 (gmt 0)

I'd still use a limited keyword tag, unique to every page.

Question: "rating" tag?

encyclo

WebmasterWorld Senior Member encyclo us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 11:00 pm on Aug 1, 2007 (gmt 0)

<meta name="dmoz.id" content="..." />

Not one I've heard of, it certainly isn't required for RSS to work, but it appears to be used to show the path to the site's category in the ODP [dmoz.org]. It is probably unnecessary in most cases.

The original poster seems to be using xhtml, as such the encoding should be defined in the very first line of the xml.
e.g.:
<?xml version="1.0" encoding="ISO-8859-1"?>

This is not a recommended way of defining the charset when serving XHTML as HTML (which 99.99%+ of XHTML is). It pushes Internet Explorer into quirks mode and is not a valid HTML method.

So if you don't include the meta charset element what is the assumed default by Google and others?

Usually ISO-8859-1, maybe US-ASCII, maybe windows-1252, maybe something else (for example the browser can attempt to guess). So don't leave it up to the user agent to make possibly false assumptions, decide what encoding you are using, declare it and use it. :)

DrDoc

WebmasterWorld Senior Member drdoc us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 11:58 pm on Aug 1, 2007 (gmt 0)

In addition to
Content-Type don't forget that should also use Content-[b]Style[/b]-Type.
klown

5+ Year Member



 
Msg#: 3409150 posted 1:41 am on Aug 2, 2007 (gmt 0)

For meta keywords google and msn don't seem to use it however yahoo does. Recently I had a site start ranking for a term that was solely inside the meta keywords (apparently its a term that nobody thought internet marketing needed to know about).

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 1:48 am on Aug 2, 2007 (gmt 0)

Question: "rating" tag?

An idealistic attempt at telling child-safe filters that your content can be trusted. I don't know of any filtering software that really uses it.

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 4:52 am on Aug 2, 2007 (gmt 0)

In addition to those mentioned above, I also utilize these on occasion...

<meta http-equiv="content-language" content="">
<link rel="start" href="http://www.example.com/sub/" title="Insert Destination Page <title> Here">
<link rel="prev" href="http://www.example.com/sub/1/">
<link rel="next" href="http://www.example.com/sub/2/">
<link rel="glossary" href="http://www.example.com/glossary/">
<link rel="help" href="http://www.example.com/help/">

As you can see, the list is fairly extensive. And, if Google keeps at it, there will be about 10 reserved for them. ;)

The above metadata elements assist in various technologies. I cannot confirm that they are of any benefit from a search engine standpoint but I do know that our /help/ files and /glossary/ sections perform very well. Note the relationship...

<link rel="glossary" href="http://www.example.com/glossary/">
<link rel="help" href="http://www.example.com/help/">

I use these all the time... :)

<!-- HTML Comments (treated as HTML markup) -->

For meta keywords google and msn don't seem to use it however yahoo does.

We've started using the meta keywords element for on-site stuff. Heck, we figure if the search engines are going to use it, we can use it for our internal search.

Even though everyone says the keywords metadata element is dead, I tend to look at it from a different perspective. We're talking about an element that has been a part of optimization from day one. While it may be very low on the list of signals, it still "may be included" in those 200 factors that Google states they look at. Many will tell you that they've tested this with some obscure keywords. Well, it doesn't work like that.

penders

WebmasterWorld Senior Member penders us a WebmasterWorld Top Contributor of All Time 5+ Year Member Top Contributors Of The Month



 
Msg#: 3409150 posted 6:43 am on Aug 2, 2007 (gmt 0)

:
<link rel="prev" href="http://www.example.com/sub/1/">
<link rel="next" href="http://www.example.com/sub/2/">
:

I agree with pageoneresults... Opera will display these links/relationships on the Navigation toolbar (if enabled). And I believe that FF (and other browsers?) will prefetch pages defined as rel="next" (or rel="prefetch") in the background - so speeding up the browser experience.
[developer.mozilla.org...]

adfree

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3409150 posted 7:45 am on Aug 2, 2007 (gmt 0)

Thanks tedster!

Fotiman

WebmasterWorld Senior Member fotiman us a WebmasterWorld Top Contributor of All Time 5+ Year Member



 
Msg#: 3409150 posted 2:43 pm on Aug 2, 2007 (gmt 0)


<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">

Why iso-8859-1? I always use utf-8, based on recommendations from W3C [w3.org].

<meta http-equiv="Content-Type" content="text/html; charset=utf-8">

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 2:47 pm on Aug 2, 2007 (gmt 0)

Why iso-8859-1?

At the time, years ago, that was the standard. When I first started designing, my default editor was generating Windows-1252. I changed to ISO-8859-1 after becoming more heavily involved in my markup.

I tried switching to UTF 8 at one time. It wasn't a simple change and I had to undo what I did. It didn't like some of the characters being used within the site and created all sorts of havoc. At the time, I didn't have the inclination to go through thousands of pages and figure out what the character issues were.

ergophobe

WebmasterWorld Administrator ergophobe us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 3:55 pm on Aug 2, 2007 (gmt 0)


>>Place this before the title element please.

It's logical - the meta charset element defines the character encoding, and so you should always define the encoding before any visible content

Yeah but... while I don't disagree in principle with the recommendation, it's a bit more involved and putting it after content doesn't necessarily cause a reparse (I believe). Though I fully agree that the tag should be used and should come first, the meta charset is not usually needed for the browser to properly parse the document.

In fact, it is actually the THIRD indicator that a user agent looks for when trying to determine character encoding. I believe that the only time this tag causes the document to be reparsed is when there is a mismatch between the encoding announced in the server header (or XML declaration if applicable) and the one in the meta tag. Using it is sort of a belt and suspenders approach, is recommended as a best practice by the W3C, but not essential (see reference 2).

In order, the user agent looks for
1.
HTTP Content-Type header sent by the server. When you set up a server, buy hosting or whatever, you should put up a test page ("Hello World") and run LiveHTTPHeaders (the Firefox extension) and find out what character encoding your server is sending. If there is a mismatch with your actual document encoding, you should change the server settings or the document encoding as a first preference, rather than overriding the setting with a META tag. It is almost universal for an Apache server to send this info (can't recall offhand if it is *always* sent or just almost always). With Apache, this can be set in httpd.conf or even in .htaccess (see reference 3 for a HOWTO).

According to the HTTP 1.1 standard, if no charset parameter is sent by the server, the encoding is supposed to be assumed to be ISO-8859-1, but in practice, because there are so many unlabelled or incorrectly labelled documents, many (all?) browsers default to the user's (i.e. browser's) preference.

2. The XML declaration. This isn't common even in XHTML because it has to precede the DOCTYPE declaration which, of course, triggers quirks mode. So though it is second in official precedence, it usually doesn't come into play.

3. meta tag charset declaration. ditto the comments above.

4. algorithmically (tries to guess based on BOM, browser preferences, whatever).

References

1. [w3.org...]
2. [webstandards.org...]
3. [w3.org...]
4. [httpd.apache.org...]
5. [httpd.apache.org...]

[edited by: ergophobe at 4:06 pm (utc) on Aug. 2, 2007]

ergophobe

WebmasterWorld Administrator ergophobe us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 3:59 pm on Aug 2, 2007 (gmt 0)


<link rel="glossary" href="http://www.example.com/glossary/">
<link rel="help" href="http://www.example.com/help/">

I use these all the time... :)

Care to elaborate? How so? What I'm asking about is the implementation. In other words, what practical purpose do these serve for the user? Or is the idea simply to have a very keyword rich link for the SEs? I understand that for the glossary, but for the help?

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 4:09 pm on Aug 2, 2007 (gmt 0)

In other words, what practical purpose do these serve for the user?

At the time I performed the research on that specific metadata, I was beginning to use Opera. When utilizing said metadata, Opera would display an additional toolbar for the user, sort of like a sub-navigation, based on those elements. It made for an "extended" user experience.

So, I figured if Opera was utilizing those elements, there must be other UAs that utilize them too. Firefox does. I'm not sure of others as I haven't kept up with who is doing what. The W3 recommends their use when applicable so I make it so. The end result is that there are added user benefits and who knows what search engine benefits, if any.

Or is the idea simply to have a very keyword rich link for the SEs?

You see, I wasn't thinking of it from a keyword rich viewpoint. I was thinking of the relationship. Glossary/Glossary, Help/Help. Those are specific relationships. And, they are "natural".

[edited by: pageoneresults at 4:11 pm (utc) on Aug. 2, 2007]

Tonearm

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3409150 posted 4:10 pm on Aug 2, 2007 (gmt 0)

When you reach a certain number of pages indexed, you want to start managing which pages are getting indexed.

pageoneresults, at about how many pages do you want to start managing the PR? Maybe it depends on the site's PR levels?

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 4:19 pm on Aug 2, 2007 (gmt 0)

pageoneresults, at about how many pages do you want to start managing the PR?

One. ;)

I probably shouldn't have used "number of pages" as a starting point. I manage everything from the first page developed.

Because of the way PageRankô works, from my perspective there are a given set of pages on a site that are not really of benefit from an indexing standpoint. I don't want to bleed off any of that PR to less important pages. Pages such as terms and conditions, privacy statements, contact us (at times), etc.

If you've only got 50 pages to work with, I would think you'd want the creme of those 50 to be at the top. One sure way to do this is not to bleed PageRankô to those pages that have little to no value to the user from a search query perspective. I really hate to use that term PageRankô because it adds that whole "design for the search engines" stigma. But, in this case, I don't want those pages in any of the search engines indices, Google is not the only SE out there you know. ;) And, because I don't want those pages indexed, one of the side effects is that they don't get PR, a natural occurrence.

Maybe it depends on the site's PR levels?

Its all about focus. Level of PR is part of it but not really the determining factor. Once you have PR, you really want to "baby it" and funnel where it is going.

ergophobe

WebmasterWorld Administrator ergophobe us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 5:05 pm on Aug 2, 2007 (gmt 0)

Care to comment on content="none" versus disallow in the robots.txt?

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 5:17 pm on Aug 2, 2007 (gmt 0)

Care to comment on content="none" versus disallow in the robots.txt?

When you specify a location in your robots.txt file, Google creates a URI only listing.

When you utilize the content="none" (robots meta tag) which is shorthand for "noindex, nofollow", Google doesn't show the page in its indice. It may index it, but it doesn't show.

ergophobe

WebmasterWorld Administrator ergophobe us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 7:00 pm on Aug 2, 2007 (gmt 0)

Ahhh! I should have asked that years ago. That explains several things I was confused about.

Tonearm

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3409150 posted 7:34 pm on Aug 2, 2007 (gmt 0)

Thanks pageoneresults. I added the content="none" tag to a few pages.

I'm using charset="UTF-8" and I hope that isn't causing any problems. Does anyone think it could?

Also, I wanted to add the Google Webmaster Tools verify-v1 tag to the big list.

edit: Have many people had luck getting Google to display their meta description tag for the home page instead of the ODP description with the noodp tag?

[edited by: Tonearm at 7:41 pm (utc) on Aug. 2, 2007]

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 7:46 pm on Aug 2, 2007 (gmt 0)

Have many people had luck getting Google to display their meta description tag for the home page instead of the ODP description with the noodp tag?

Yes, in several cases. Google does honor it.

g1smd

WebmasterWorld Senior Member g1smd us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 7:52 pm on Aug 2, 2007 (gmt 0)

>> the meta charset is not usually needed for the browser to properly parse the document. <<

Just to note that the character set for the document must be declared somewhere otherwise the document is not a valid HTML document.

Tonearm

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3409150 posted 8:37 pm on Aug 2, 2007 (gmt 0)

Thanks tedster, does the noodp tag need to be on every page or just the home page?

BTW, does a meta description need to be a complete sentence or can it be a series of keywords?

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3409150 posted 9:04 pm on Aug 2, 2007 (gmt 0)

Meta descriptions should be grammatical, although partial sentences are OK, such as "The most complete widget collection on the web."

Keyword lists are not a good idea for a meta description. They would be unlikely to get used as the snippet anyway -- and if they do, they're less likely to draw a click.

Don't get too tricky here, looking for a significant ranking boost to come from your meta description tag. That's not too likely, and trickiness can come back to bite you. Just use the meta description as a place to put an attractive and uniquely relevant description of that particular page.

Most search engines truncate the snippet at around 160 characters today or even less, so keep the meta description short. Think of it as a short marketing blurb for the page. It's not a direct way to raise your rankings -- except that your rankings may go up when your description draws a lot of clicks whenever it gets an impression in the SERP.

The noodp meta tag should be on any url where there is an ODP entry for that url. This may be just the home page, or it may include subdirectores and subdomains, if they also have an ODP entry. I rarely use the noodp meta unless I see that a search engine is choosing the odp description. When page titles are well-written and uniquely specific to each page, that problem doesn't usually happen unless there is an extended spidering problem.

Page titles and meta descriptions deserve at least as much focused consideration as the on-page content gets. When they are well crafted, they create a free ad for the page in the search engines.

[edited by: tedster at 12:44 am (utc) on Aug. 3, 2007]

This 55 message thread spans 2 pages: 55 ( [1] 2 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Code, Content, and Presentation / HTML
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved