So, Meta Tags Are Dead ?
#1
Posted 27 August 2007 - 12:06 AM
I've decided to do research into SEO - I thought I had the basics... and my sites do okay... but that isn't enough... they have to do well - really well!
So I read. I look up. I research. I join Google and ask questions...
...at which point I refer to my foul terminologies...
I can accept that there is no "hard and fast" for SEO... but I expected at least some "known and quantifiable" information. Instead, it seems that a large part of SEO is based on guess work.
Most confusing of all... I am told/informed/finding out time and again that tings such as the Description and the Keywords haev virtually no real value when calculating a pages SER value - it's not like it was!
Is this true?
If so - what the hell are they used for?
If so - are they worth including?
If so - do they cause any negatives by including them?
Please note... I am not refering to the pointless green bar (as for some reason whenever I mention page rank or Search Engine Rank - people automatically conclude I'm refering to that silly bar!).
I am talking about getting ito the first page... the top 10 for things like Google, MSN and Yahoo.
So answers - and if at all possible... links to something offical stating the answer.
I'm so tired of hearing contradictions ad opinions - is there no hard facts left for this sort of thing?
#2
Posted 27 August 2007 - 01:22 AM
Welcome to the forum!
Okay...here's the thing. There are only a couple of guys at Google who actually know Google's algorithm.
This means that, yes, SEO is based on what you might call guesswork, but which we prefer to think of as experience and education. SEOs with years of experience watching Google react to different elements, pages, sites etc. are able to make pretty decent educated guesses as to what Google places importance on.
You have heard correctly that the meta keywords tag is, in effect, dead. Forget about it.
The meta description tag also will not influence your rankings. HOWEVER, if you do use a meta description tag, Google will often (not always) use this as the second, descriptive line in your search engine results. The standard format is that the 1st line is your Title Tag and the 2nd line is your meta description. Google may also decide to pull the meta description from elsewhere (such as DMOZ), or if you leave it blank, Google may decide to pull a description out of the main text of your page.
If you do choose to have a unique meta description for each page, it should act as a marketing pitch. It should be a line or two of text that would make me so excited about your listing that I want to click on it instead of on the listing of your competitor. This doesn't mean saying "we're the best". It's more like:
An in-depth account of Medieval hairstyles, including full color photos, illustrations and do-it-yourself instructions.
If I'm looking for information on Medieval hairstyles and read this in your listing, you look pretty compelling to me and I want to click.
So, that's the point of meta descriptions...not for rankings, but for clickthroughs.
There are no official documents, however. Search engines don't work like that. However, there are Google's official webmaster guidelines which give you a fair, if broad, idea of what they want and don't want. For example, Google doesn't want siteowners to buy links or engage in cloaking. They do officially state this. But they don't officially tell you how to be #1, and neither will any legitimate SEO. It doesn't work that way.
You do need to learn technical basics (like what is a title tag, a meta tag, etc.) but once you have those straight in your mind, you will quickly learn that your main focus needs to be on delivering valuable, linkworthy text content to your public. I know it can be frustrating at first, with so many different sources telling you different things. You can take my word for what I'm saying, or you can find it out the long way by seeing how search engines treat different things you publish. That puts you in the same boat with the rest of us, watching and studying how search engines work.
My only other suggestion is that you begin studying Google patent applications, if you like official documentation. Bill Slawski of http://www.seobythesea.com has spent countless hours writing about patents Google has applied for. Studying these can provide unique insights into potential factors in Google's algorithm, but even here, there is guesswork going on.
Stick with it!
Miriam
#3
Posted 27 August 2007 - 07:45 AM
have a look at http://www.choose-easyweb.com - if you view the source you will see that I matches within;
<URL> = Sitename + PageTitle (+ParentPageTitle)
<title> = SiteName + PageTitle
<keywords> = SiteName + PageTitle + Keywords (Not all are mine/legitimate to content!)
<Description> = SiteName + PageTitle + Description (Inc. certain Keywords!)
<h1> = Sitename + PageTitle
<h3> = Page Title
<p> = Content =>consisting of PageTitle, Links with partial matching Titles, several Keywords and partial match to Descirption
All of that... only to find the Keywords/Description basically do naff all.
Worse still - apparently having "good matching" can actually result in penalties being applied (which strikes me as rather foolish - it would only be logical to have the Page Title repeated throught out, and the Page Title should berefered too in the Description and Keywords!)
So - is all of that worthless now?
Or is it simply we cannot really tell what it is worth - but we can have a guess?
Any insight or comment is more than appreciated.
#4
Posted 27 August 2007 - 08:23 AM
Next we have two name values: keywords, which these days is mostly useless, ironically, and description, which is still somewhat useful.
Treat them like believing in the 'force'. Good if it is true but causes no harm if you wrong!
Yannis
#5
Posted 27 August 2007 - 10:18 AM
In fact, the algorithm isn't that simple. Indeed, when people talk about 'the algorithm' it is misleading. There isn't an algorithm really. There are algorithms - plural, multiple, many. Each factor that can be mathematically measured has its own algorithm, and each algorithm relates to others, some with more weight, to create an overall measure of relevance.
A great many of those factors are not based in the HTML. One of the largest shifts in the latter 90s was into using off-the-page factors, most notably weighted link popularity, as a more independant measure of a page's true merit and topic. Soon they also added time-based measures, so that fresher content received a little boost to compensate for the fact it would have less links (helping news and updated info rise above old favourites for a while).
If you think how complex your own mind's mental calculations are about the quality of a page, and all the clues and measures it uses to assess how trustworthy and valuable a page is, then I'm sure you see how low meta tags rate in that equation for you, the user. And that is what a search engine is trying to emulate.
The keywords meta tag proved virtually useless as any kind of data signal. Not only was it often deliberately abused, stuffed full of barely relevant words, along with irrelevant ones, and with no impartiality. There are also thousands of websites that use the precise same keywords for every page of their site, meaning these tags do absolutely nothing to differentiate one page from another. You can honestly leave the keywords meta tags out of a page and see no loss of ranking power at all.
The Description meta tag is far more useful though. It won't help you much with actually getting ranked high. But once you do get a high ranking with all of the other factors, a good description can be the make or break of click-through percentages.
Yet the Description is not always the snippet shown in the SERPs. That started with a search engine called FAST, which later was incorporated into the overall search technologies of Yahoo! FAST felt that with all those sites with the exact same description on each page, it was as often worthless as not. They instead chose to use the first visible text in the body of the web page.
It was awful, because the first text on many pages was the top navigation links like login and register and search, etc. Made for a lousey and useless description for the poor searcher.
Another method for getting more 'reliable' descriptions was when the search engines looked for a matching description in human edited directories. For homepages mostly, you'd see the DMOZ description used either alongside, or instead of, the meta description. Fairer in the main, but sometimes those DMOZ descriptions were very poor outside of the context of the category structure of the directory.
What Google does these days is look at your meta description and see if it uses the same words early on as the search query used to return it as a result. If there's a match, then it will use your meta description as the snippet. If not a match, on the case by case search query, then it uses the first match from the page instead and returns the text around that as the snippet.
So you can still use a meta description to boost click-throughs from the SERPs if you understand which queries will likely return that page in the results, and use those words in your description.
But no meta tag is really going to help you much with getting those rankings in the first place. When you think about that, and how rarely actual people look at the meta tags while using the web, and hoping a search engine will help them find what they seek, well, it makes perfect sense, and does improve results.
#6
Posted 27 August 2007 - 01:48 PM
The holy grail of SEO does not exist. If it did, you would have found it by now. All you will find are opinions, nothing more. Some of these opinions will conflict with one another, even amongst highly respected and well-known industry figures. What is one to conclude when this happens? That is entirely up to you. Many will try an experiment to form your own conclusions.
The search engines have spent a considerable amount of time, money and effort on continually striving to perfect their algorithms. I've read there could be as many as 300 different factors they consider. Google dominates the market share with their exceptional search engine which I would speculate is worth many billions of dollars. It would be naive to think they (or any other search company) would actually publish anything that would help anyone to figure out how it actually works.
Patents give you 'glimpse' into some thoughts behind one of the components. However (and I say this with no disrespect to our resident patent expert), it is but only a glimpse and does not help you to see the entire picture clearly. It's analagous to saying KFC uses cayanne pepper in their famous chicken recipe.
You can jump up and down all you want, but you won't find the authoritative document you're looking for -- if you do, let us know where to find it.
Edited by Respree, 28 August 2007 - 08:22 AM.
#8
Posted 28 August 2007 - 01:04 AM
Do the basics -- the things most people agree on and hope for the best. Focus on creating the best content you possible can and, if you're successful in this regard, you'll find people will think you're a pretty good resource and naturally link to you. When they do, you'll find things will be going in the direction you want them to.
Good luck -- to us all.
#9
Posted 28 August 2007 - 01:48 AM
Oh don't get us wrong. It is logical. It just isn't simple.It's stupid really - by nature I'm a cynic...but for some reason I honest though the performing SEOwas a logical process.
Medicine is logical, and has been practiced for centuries, or indeed millennia in its basic forms. Yet there are still so many many things that are not fully understood, and so many things that escape simple diagnosis.
The search engines employ hundreds of brilliant PhDs, and they don't do that so they can all sit on their thumbs praising the one simple algoithm of the founder. Instead, they are all in little teams, each working on perfecting yet another intricate part of creating a machine that has enough AI to make judgements about quality, to guess what people really mean when they give a three word search phrase, and to try to best supply the right result for each query.
They still aren't anywhere that close to that ultimate goal, but you can be sure t is a very very complex mechanism, made by hundreds of the best in their field, and too complex for any one of them to really grasp the inticacies of exact detail of all the diverse parts.
Does that help explain the fact nobody can just write out a list of ten simple steps to a number one ranking? Not even the guys who build the algorithms? Its not a lack of logic, just an overdose of complexity of many logical processes. There are simply too many diverse variables across all of those different parts. Each logical, but each dependant on others to build up an entire picture.
#10
Posted 28 August 2007 - 03:55 AM
Patents give you 'glimpse' into some thoughts behind one of the components. However (and I say this with no disrespect to our resident patent expert), it is but only a glimpse and does not help you to see the entire picture clearly. It's analagous to saying KFC uses cayanne pepper in their famous chicken recipe.
Cayenne. I wondered about that.
More than the processes described in the patent documents from the search engines, what I tend to look for are the ideas and assumptions behind them.
For instance, what does the growth of the use of mobile devices to connect to the Web mean to the way that a search engine might look at pages. How might they attempt to distinquish between Denzel Washington and Redmond, Washington? Is there value to letting customers write product reviews on the pages of your ecommerce site, and what might happen to those reviews if they do?
Doing the basics is a good start. It's like building a solid foundation for your site to work upon. But there are other things to look at, too, such as making it as easy for people to use your site as possible once they arrive there, and using the words that your intended audiences expect to see (which helps with search engines, usability, and conversions).
[corrected typo in quote]
Edited by Respree, 28 August 2007 - 08:24 AM.
#11
Posted 04 September 2007 - 04:35 AM
The fact is, you can analyze a couple hundred thousand search results
and query the data. But you don't get facts.
You just get "trends". Like everyone above had said, the algorithm
is a bucket of worms. Still, some correlations seem to exist and
picking the best of the correlations works for my sites and my clients sites.
-The meta-keyword borders on being harmful to rankings.
Some months it is and some months not.
-The title is the single most powerful word(s) on your page.
Don't waste it on unsearched words.
-The meta-description has a correlation of ....let me go check....
Data: [+78] 50 41 38 39 39 38 34 28
There ya go.... +78.
That's not 100, but 78 is a valid correlation.
Use your desired keyword in your description tag.
Experiment with 2-3 times.
That about covers it.
The Holy Grail of SEO.
Google printed some guidelines and everything they
say in those guidelines matches the data I use to rank sites.
It preaches " Unique Content" and they are changing the
system more and more so anyone can...
Rank On Google @ #1
by SEO Expert
SEO is (now) simple as pie.
CONTENT is king. You need know NOTHING about "SEO."
LSI = Write on-topic
Two extra pages of text (on-Topic) can trump any secret and
beat any competition. ANY competition. Any Secret.
( except for that 301 redirect that just causes people harm)
Add another page of relevant text to your page and it will rank higher.
Not a new page. Add 500 more words of relevant text to your EXISTING page to get it to rank higher. Then do it again. Don't stop.
There is no penalty for too much relevant content on a page.(tip)
Before you object.....do it.
My clients pay me
to show them
that they can do it all themselves
without my help.
Your spending time here
when you could be writing.
Edited by ChuckTM, 04 September 2007 - 02:42 PM.
#12
Posted 04 September 2007 - 08:10 AM
Yet, I hate to say this,it sounds as if the contet should be aimed at the Search engies.
I thought the idea wasto haev content for "humans", and then tweak the site code to tie-in with it to et better SER.
???
So I write, naturally, about red wellington boots.
I have a frontpage introducig the general site topic.
URL -> redwellies.com
<Title> of -> Red Wellies - Welcome to Red Wellies
<M-Desc.> of -> Red Wellies - Welcome to Red Wellies - Your one stop shop for information about Wellies, Wellingtons, Waders and other water proof boots, shoes and footwear.
<M-Desc.> of -> Red Wellies, Welcome to Red Wellies, information, Wellies, Wellingtons, Waders, water proof boots, water proof shoes, water proof footwear.
<h1> of -> Welcome to Red Wellies
<body Ps> -> Several intro paragraphs.
P1> Wellies & Wellingtons & Red Wellies
P2> Wellies & Waders & Waterproof
P3> Materials & Waterproof & Sizes & Length +
Then there are multiple pages, followig a similar pattern, focusing on each Section...
Materials -> Rubber, Plastic, Poly's, Waterproofed Canvas etc.
Sizes -> Small, Large, Calf, Knee, Thigh, Lower body
Design -> Plain, Boring, Traditional, Flashy, Elegant, Fancy
Know... to me, that is how it "should" work... I write about something important... then create the Tags that match.
It jsut really does seem that the issue is "Keywords".
People automatically think of the Meta-Keywords
... not ...
"keywords - those expressingthe main focal point of the topic of page cotent and/or sitecontent".
#13
Posted 04 September 2007 - 10:51 AM
i'm simply curious as to what you believe this holy grail is. i'm more skeptical about it myself because i believe that once you find an optimization strategy that works, stick with it but i don't think there's a full and perfect formula for seo.
maybe i'm just naive or something.
-tambre
ps. sorry if this is horribly off topic or anything
#14
Posted 04 September 2007 - 12:05 PM
This is nice, except that getting links is really sort of viral... human-powered viral. ;-)-The meta-keyword borders on being harmful to rankings.
Some months it is and some months not.
-The title is the single most powerful word(s) on your page.
Don't waste it on unsearched words.
-The meta-description has a correlation of ....let me go check....
Data: [+78] 50 41 38 39 39 38 34 28
There ya go.... +78.
That's not 100, but 78 is a valid correlation.
Use your desired keyword in your description tag.
Experiment with 2-3 times.
Oops. That really is off-topic.
Edited by AbleReach, 04 September 2007 - 12:09 PM.
#15
Posted 04 September 2007 - 02:24 PM
So... we all aknowledge that Meta-Keywords are a more than a little redundant.
So what does that leave us with?
Title
Headers
Content...Important
Links in to your site... important
anything else?
Alt Text
Link Title text
????
others?
#16
Posted 04 September 2007 - 02:31 PM
2. Content...Important
3. Links in to your site... important
Ya, DON'T try to rank for 7 keywords per page!
Try one for starters. (secret insider tip)
You can always add more when you get to #1
for that keyword.
Keywords
The issue USED to be keywords. That led to Keyword stuffing.
So Google has fixed that flaw. Now the keyword can still be measured
and found, but NOW is must be surrounded by its (LSI) FAMILY of words
to be relevant. So if you were to take some random story and sprinkled
the keyword on it, it no longer ranks well.
Now - you should have the keyword in its correct family.
An idea so cool, they made a tool:
https://adwords.goog...ordToolExternal
That's something that happens if you know what you talking
about, and doesn't if your writing when you don't know Jack.
Like sooo many of those adsense sites out there.
Edited by ChuckTM, 04 September 2007 - 02:36 PM.
#17
Posted 04 September 2007 - 09:53 PM
IMHO the SERP magic happens when people, ideas, needs connect and spark off of each other. Someone likes your site or needs what you know or blips into your site when the site is ready to be remembered. Content grows in response to your content. And when links happen, the map, the phone book or whatever will trigger more and bigger blips on the radar, causing money-making traffic numbers to flow.
Reply to this topic

0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users






