|Developing Effective META Description Elements|
The importance of this element has never diminished!
Building the Perfect Page - Part III - The Basics
Developing an effective META Description Element
I wanted to resurrect a very basic discussion concerning the META Description Element. I'm finding way too many instances of sites experiencing problems which may be due to issues revolving around meta descriptions, duplication, and believe it or not, lack of having one.
First, let me expose one of the largest drawbacks of using a meta description...
Duplication of meta descriptions may cause issues. In fact, it has been proven to cause issues and there are quite a few topics here at WebmasterWorld which discuss those issues. One being, that pages that had near and/or duplicate meta descriptions were being dumped into the Supplemental Index (when we could see it).
I've seen discussion over the years where people have claimed that the meta description had little value, I think they were blinded by the fact that "they couldn't SEE" what their meta descriptions were doing.
I look at meta descriptions as the "meat of the page". I use an Inverted Pyramid Writing style approach or what is also referred to as Front-Loading.
|Note. This is commonly referred to as "front-loading" and is especially helpful for people accessing information with serial devices such as speech synthesizers. |
13.8 Place distinguishing information at the beginning of headings, paragraphs, lists, etc. [Priority 3]
If you are serving dynamic meta descriptions provided from a feed of some sort, you may have duplication issues to contend with. If everybody (affiliates, etc.) are using the same meta descriptions, I think that would pretty much negate the use of that element within that segment? The site with the highest PR will usually win.
So, how are you handling the development of your meta descriptions?
If you get a feed, are you doing anything special to "break the chain" and address the duplication issues?
How important do you think the META Description Element is?
For Google and Msn seaches, it is used to describe your site. It is there to help the searcher make a more informed choice on which link to click. I think its critically important for users who dont blindly click on the first result or have not found what they wanted on the first result. Two well written sentences should summarize your site and increase traffic.
Now a question. How can you optimize your site description in Yahoo searches? Does it completely ignore the meta description?
|Now a question. How can you optimize your site description in Yahoo searches? Does it completely ignore the meta description? |
Good question, let's see if I can offer "my personal insight" on this. Yahoo! does utilize parts of the meta description in its snippet generation. It also uses content that it finds after the <body> element. So, this means that whatever is sitting after the <body> after the html is stripped, that is relevant, is a likely candidate for other parts of the snippet. It could be parts of your top and left navigation and in many instances will be if the search phrase appears there and that is what sits right after the <body> element.
This is where I feel strongly that SOC (Source Ordered Content) comes into play, some will disagree with me. But, after years of working with SOC, I'm convinced it does play a "major role" in defining the "meaning" of your pages.
So, you have this page sitting there and it is Front-Loaded (including the meta description) and utilizes Source Ordered Content and an Inverted Pyramid Writing style approach. Based on what you are seeing in the results returned in Yahoo!, how do you think that would affect those results?
My experience with Google is exactly the same - duplicate or missing meta descriptions have been a primary reason for otherwise good pages being dumped into the supplemental index.
I know it takes extra effort to make sure these meta description elements are both present and unique -- especially on dynamic pages. With some CMS systems, it can take a custom modification to make it even possible! But the advantages are major, and I am intense with clients about this area. It's not just a nice thing - it's one of the very top "best practices" today, in my view.
[edited by: tedster at 12:02 am (utc) on Aug. 31, 2007]
What is duplicate meta description?
If I have meta description = "Blue widgets are blue, Green widgets are green."
and in my body tag I have the introductory paragraph "Blue widgets are blue, Green widgets are green.".... is this considered duplication?
I'm asking because I usually copy my introductory paragraph and paste it as meta description.
Interesting, as the (sort of) consensus I was hearing was that having no meta description was better than having a generic one...
Yes. This is a factor. There are several sides to this.
One of them is where multiple pages within a site bear the same title and/or meta description. Matt Cutts confirmed it just a year ago, over there: [threadwatch.org...]
The other is where multiple sites have the same title and/or meta description. I have seen this at work first hand. I helped a site owner clean up their site, and improve the code, content, internal navigation, and meta data (and titles). The site did a lot better in search results within weeks.
Some months later, I had the opportunity to write their one-page entry for the "directory" section of the national association website of whatever it was they did. The previous entry just listed the business and their address and telephone number. The new entry included a link to their website, email addresses, and a couple of paragraphs about what the business did.
One thing that I did (unintentionally), was to simply copy the meta data from the "about us" page on the site, to the entry page on the association website. Within days, that association page was #2 in the SERPs, just below the real site. Both titles and both snippets were showing identical data. Then, a week or two later, one of the pages simply vanished from the SERPs (the association page).
I wrote this up at the time. It was one of the early clues as to "Duplicate Content" filtering, several years ago. I recently tested this again, using a site that wasn't worried if one page of their site dropped out of the SERPs for a while. I got broadly similar results.
I look after many websites and have most sites with meta descriptions included, but as an experiment I deliberately left out some meta descriptions and keywords - conclusion they still got ranked - don't know whether this changes as a result of strong competition in the SERPS
From my standpoint I couldn't agree with pageoneresults or tedster more. Most everyone of the pages we make utilize the META description and keywords attribute. Additionally, just as PowerUp describes we use the opening text(in large part) for our META description.
Over the past 10+ years the one thing I've learned is that it's the relevant quality of the data on the page which will really help your SERP position. Not that we're #1 for akk keyword phrases but we try our best.
So, I would agree that META description Elements are vitally important to page development/success, even though it's importance/relevance is being down played by so many. Personally, it is important because it's an indication that someone has taken the time to analyze each page/product/sku and write relevant copy and that someone is will to do that "little bit extra".
So, who is willing to do that "little bit extra"? I am, are you? :)
When I write articles for my site I have a field for META description where I can put anything I want. I usually put in a one sentence explanation/overview of the article and try to make it sound catchy without overselling it. Is one sentence enough or would you recommend two sentences or more?
My pages seem to rank well using my current one sentence descriptions but if they can rank better then I am willing to put in some extra time to update and lengthen the descriptions.
|What is duplicate meta description? If I have meta description = "Blue widgets are blue, Green widgets are green." and in my body tag I have the introductory paragraph "Blue widgets are blue, Green widgets are green."... is this considered duplication? |
From my perspective, not at all. I'm used to "Front-Loading" the first paragraph so it makes a great candidate for a cut and paste routine. But, I will also make slight modifications to those meta descriptions so that I can confirm that what I'm seeing in the SERPs is a replica of the meta description and not the first <p></p> on the page. ;)
|Interesting, as the (sort of) consensus I was hearing was that having no meta description was better than having a generic one. |
In the above instance, then it is up to the SE to determine what the page is about based on the content that comes directly after the <body> element. Which in most instances is not enough. That's why I think it is very important that the meta description element mirror what is in those first two paragraphs. Sometimes I'll pull the first sentence from the first <p>. Then the first sentence of the second <p>. I may even include an opening sentence that uses words from the <title> and <h1> to help "bring the page together".
There's a whole bunch of "handshaking" going on in the <head> and that content that immediately follows the <body> element. :)
|Is one sentence enough or would you recommend two sentences or more? |
I usually use one or two sentences. Typically around 160-180 characters although that is not a hard rule. If I can describe the content of the page concisely in one sentence, then so be it. It's probably going to be a looong sentence. :)
Meta description: I often use about 15 to 25 words as a sentence.
I'm going to put the Wiki on the spot here and use them as an example of "what not to do"...
Take a look at the SERPs. Expand your results to 100. Now take a close look at the snippet generation. Is that what you want to leave up to chance by not having a meta description element?
In the Wiki's case, the snippet is being generated based on the content directly after the <body> element.
|From Wikipedia, the free encyclopedia. Jump to: navigation, search. logo. |
What a waste of snippet space right there. The insertion of a meta description element that concisely describes the page would do wonders for the Wiki. Not that they need much more help. :)
Some mega-sites can get away with that.
Your "average" site with a few dozen to a few thousand pages, cannot.
OK, but in theory while the snippet generation may not read well, it would also be less likely to raise the duplication/ supplemental problem of using a generic description.
|OK, but in theory while the snippet generation may not read well, it would also be less likely to raise the duplication/ supplemental problem of using a generic description. |
But the Wiki would need not worry about that. They have enough unique content in their first paragraphs to generate an optimized meta description element.
|But in theory while the snippet generation may not read well. |
In the Wiki's case, they can get away with it since they are pulling authority positions above the fold. They are relying on the page <title> to be the "clicking factor".
Me? I'd prefer to light up that listing as much as possible. <title>, meta description and URI. Target and optimize all three and viola, instant "click appeal".
clicking factor + click appeal + URI_KW =?
what does it to your STATS?
META Description= H1 + "p1(3 main targeted words)+li2+li3+li4"
try it, long tail!
|My experience with Google is exactly the same - duplicate or missing meta descriptions have been a primary reason for otherwise good pages being dumped into the supplemental index. |
Boy, is this true. I learned it the hard way, when I used almost identical meta descriptions when generating about a hundred pages for a directory.
It went like this:
Repairs and service for widget rutabagas in Smallville, Texas
Repairs and service for widget rutabagas in Anywhere, Texas
Repairs and ... (you get the idea)
Ranked fine for about 6 months, then went deep, deep into supplemental territory.
Don't do this! It hurts!
--Repairs and service for widget rutabagas in [Anywhere], Texas--
Could be a lot of "where" there, where "there" is a combination of linked in by other sub pages with juice that is just compensated by whatís on the page linked from and on page it self.
I know it is not what seems to be a "shoot the ghrabit" situation, just an observation...
blend, thanks for the comment, but I'm totally in the dark as to what you said. Could you try again? Sorry for my being so dense.
--- shoot the ghrabit Ė
If you slice it to the extent that the linking pages, in content, have enough juice for the 'shooting in the dark of Smallville township' with the combination on another page where 'Smallville Rabbits shooting blanks', the 'Dark side of rabbits from Smallville shotting blanks' suddenly becomes relevant to both queries prior.
The benefit of the page linking to the page in question and the percentage of the link text variations pointing to the next page should(donít shoot the rabbit) semantically reframe, topic, from with in or a similar sites/pages.
At the end, it's only text.. related.
look from outside-in, what would describe your best link in your description tag? your wish..
|It went like this: |
Repairs and service for widget rutabagas in Smallville, Texas
Repairs and service for widget rutabagas in Anywhere, Texas
That may be considered "near duplication" where there are only one or two words changing within a generic meta description.
Google, Yahoo!, Live and Ask all utilize the meta description. Google tends to use the full meta description in more instances than the others. The others appear to grab the first sentence of the meta description and then something from right after the <body> element. And that may not necessarily be to your advantage depending on your page layout.
Working with dynamic meta descriptions is tough. It takes strategic planning to piece together a meta description that reads well and best represents the content on the page. The other thing you have to watch out for is the duplication issue. The more variables you are working with, the less chance of crossing that duplication threshold. Using just one or two variables in the meta description may not be enough. We've sometimes got upwards of 10 variables depending on the page content and what is available to us.
For example, let's say I have a directory of local business listings. I may include the address at the end of the meta description. Street Address, City, State, Zip, Telephone Number, etc.
I've been under a rock for a while, so excuse me, but...
Let me try my hand and see if I got it:
|clicking factor + click appeal + URI_KW =? |
Are you asking a question? Rhetorical?
|META Description= H1 + "p1(3 main targeted words)+li2+li3+li4" |
Are you suggesting this is a good formula for dynamic META Description? The page's H1 (main heading) + 3 main targeted words from the 1st paragraph + the 2nd, 3rd, and 4th list items from perhaps the Top Nav or Left Nav?
"Long Tail" - I've heard this a few times recently. What does that mean?
Thanks for helping me get back up to speed and thanks for another great post pageoneresults.
Long tail: all those searches that are obscure and which bring you one visitor per month but when added together account for more than half of your traffic.
--META Description= H1 + "p1(3 main targeted words)+li2+li3+li4"--
I wrote a CMS that uses this(and few other :)) formula
H1 is H1
p1(3 main targeted words) would be the words that you are targetting to rank for long tail from the first par on the page.
li2+li3+li4 are list items that describe the widget properties and/or funk'tions
we have a large inventory of li', so it is dynamic. there are separators between the members of the formula, and I gotta tell you that once in a while the generated DESCRIPTION does not make sence, and that page does not rank even for its own H1 in quotes, at least on G, for Y it does not matter as much as it does not look at description tag as much.
It just make sence, and the coolest part is that I learned how to structure the pages that rank HERE, at WebmasterWorld.com!
Great topic - thanks pageoneresults.
Often, the lowest hanging fruit for an old site that has never had any seo applied to it are the metas. One of the ways I work the meta description is to place the most important keyword at the beginning.
Google tends to pick up the meta description for the most important keyword searched in my experience. So, I counted the maximum character length typically shown in the SERPS, usually around 154. So my first 154 characters are the most important, and I use compelling language combined with keywords to hopefully get the click.
Let me tie this together: With a Google PPC ad, you only have 25/35/35 characters per line for your ad. Thinking of your metas as an ad to help compel the searcher to click on your result usually yields great results.
Google displays roughly 66 characters for the Title, and 154 characters for the description. By optimizing these two metas with the most compelling, kw rich text, within this character limit, it's like writing a longer advertisement. You can have longer metas, but they won't display.
That's a great observation, Justablink.
I am going to take a look at that first 154 characters in my metas. Seems worth it to remove anything extraneous like and's The's etc.
I believe a description tag thats longer, in order to include a few long tail kw's, is still worth it... just won't display.
I've considered meta descriptions important for a long time simply because they are often used as the snippet in search engine results.
But are some of you saying there is something more to them?
Is it possible that Google or other SEs consider that abbreviated summery of what the page is about in their algo?
Or is it more a matter of the SEs seeing the first few words on a page as important in telling what a page is about and if there is a meta that simply comes first after the page title?