Meta tags are the beginning of most SEO training, for better and worse. I contemplated on how to introduce this topic because we always hear the bad part of meta tags, namely the keywords meta tag. One of the first things dissected in any site review is the misuse of meta tags, mainly because they are at the top of every page in the header, therefore the first thing seen. But we don't want to get too negative, meta tags are some of the best tools in a search marketer's repertoire.
There are more meta tags than just description and keywords, though those two are picked on the most. I've broken down the most used (in my experience) by the good, the bad and the indifferent. You'll notice that the list gets longer as we get to the bad ones. I didn't get to cover all of the meta tags possible to add, but there is a great Meta Tag resource you should check out if you're interested in what is out there.
My main piece of advice is to stick to the core minimum, don't add meta tags you don't need as they just take up code space. The less code you have the better. Think about it like this, your page code is like a set of step by step directions to get somewhere, but for a browser. Extraneous meta tags are the annoying 200 feet line items in directions that tell you to stay on the same road you were on!
The Good Meta Tags
These are the meta tags that should be on every page, no matter what. Notice that this is a small list, these are the only two that are required, so if you can work with just these two, please do.
- Meta Content Type - This tag is necessary to declare your character set for the page and should be present on every page. Leaving this out could impact how your page renders in the browser. A few options are listed below, but your web designer should know what is best for your site.
- Meta Description - The infamous meta description tag is used for one major purpose, to describe the page to searchers as they read through SERPs. This tag does not influence ranking, but is very important regardless. It is your ad copy that will determine if the user clicks on your result. Keep it within 160 characters, and write it to get the user's attention. Sell the page, get them to click on the result. Rand wrote a great article in 2007 on meta descriptions that goes more in detail.
Edited In: I left out the title tag as it doesn't "start" with meta, but it is still technically a meta tag. You should always have a unique title tag on every page that describes the page. Check out this post for more information on title tags.
The Indifferent
Different sites will need to use these in specific circumstances, but if you can go without, please do.
- Robots - One of the largest misconceptions is that you have to have a robots meta tag. Let's make this clear: In terms of indexing and link following, if you don't specify a meta robots tag, they read that as index,follow. It is only if you want to change one of those two commands that you need to add meta robots. Therefore, if you want to noindex but follow the links on the page, you would add the following tag with only the noindex, as the follow is implied. Only change what you want different than the norm.
- Specific Bots (Googlebot) - These tags are used to give a specific bot instructions like noodp (forcing them not to use your DMOZ listing information) and noydir (same, instead the Yahoo Directory listing information). Generally the search engines are really good at this kind of thing, but if you think you need it, feel free. There have been some cases I've seen where it's necessary, but if you must consider using the overall robots tag listed above.
- Language - The only reason to use this tag is if you are moving internationally and need to declare the main language used on the page. Check out this meta languages resource for a full list of languages you can declare.
- Geo - These meta tags, last I heard, are supported by Bing, but not Google (you can target to country inside Webmaster Tools). [Updated: Matt Cutts speaks of the tag here, they don't really look at it still] There are three kinds: placename, position (latitude and longitude) and region.
- Source - The new kid on the block, this is a tag (really two tags) that Google is testing out (thanks to the tip from my coworker Justin Briggs). It's mainly for news providers so that they can help the search engines understand who the original news source is and which ones are syndicates. The news world and search world are trying so very hard to get along, glad to see this one pop up.
- Keywords - Yes, I put this on the indifferent because while no good SEO is going to recommend spending time on this tag, there is the small possibility it could help you somewhere. So please leave it out if you're building a site, but if it's automated there is no reason to take it out.
- Refresh - This is the poor man's redirect and really should not be used if at all possible. You should always use a server side 301 redirect. But I know that sometimes things need to happen now. But Google is NOT a fan.
- Site Verification - Your site is verified with Google and Bing right? Who has the verification meta tags on their homepage? These are sometimes necessary because you can't get the other forms of site verification loaded, but if at all possible try to verify another way. Google allows you to verify by DNS, external file, or by linking your Google Analytics account. Bing still only allows by XML file or meta tag, so go with the file if you can.
The Bad Meta Tags
There is nothing that will happen to your site if you use these, let me just make that clear. They are a waste of space though, even Google says so (and that was 5-6 years ago!). If you're ready and willing, it might be time for some spring cleaning of your area.
- Author/Web Author - This tag is used to name the author of the page. It's just not necessary on the page.
- Revisit After - This meta tag is a command to the robots to return to a page after a specific period of time. It's not followed by any major search engine.
- Rating - This tag is used to denote the maturity rating of content. I wrote a post about how to tag a page with adult images using a very confusing system that has since been updated (See the comments). It seems as if the best way to note bad images is to place them on a separate directory from other images on your site and alert Google.
- Expiration/Date - Expiration is used to note when the page expires, and date is the date the page was made. Are any of your pages going to expire? Just remove them if they are (but please don't, keep updating content, even contests, make it an annual contest!). And for date, make an XML sitemap and keep it up to date, that is so much more useful!
- Copyright - That Google article debates with me here, but look at the footer of your site. I would guess it says "Copyright 20xx" in some form. Why say it twice?
- Abstract - This tag is sometimes used to place an abstract of the content and used mainly by educational pursuits.
- Distribution - The distribution value is supposedly used to control who can access the document, typically set to global. It's inherent that if the page is open (not password protected like on an intranet) that it is for the world. Go with it, and leave the tag off the page.
- Generator - This is used to note what program created the page. Like author, useless.
- Cache Control - This tag is set in hopes of controlling when and how often a page is cached in the browser. It's best to do this in the HTTP Header.
- Resource Type - This is used to name the type of resource the page is, like "document." Save yourself time, as the DTD declaration does it for you.
