Forum Moderators: open
Is this advisable or will Google penalise him?
Steve
Msg 17 by Tedster :-
The h1 tag is [supposed to be] the heading or topic of the entire page. I've seen several mentions on the W3C that having more than one <h1> is considered an error. Although it's not invalid code, strictly speaking, it is a poorly structured document.The W3C is creating an accessibility checker -- multiple h1 tags are one of the errors that it will flag, along with skipping from <Hn> to <Hn+2>, instead of following <Hn> with <Hn+1>
I think discussions like these about 'how many times can I use ...' are no longer a priority. A keyword density of 8% could have killed you a year ago, now you can go as high as you like .... alt tags, headers, comment tags, title tags, bold/italic/underlined text and so on and so on don't seem to make the slightest bit of difference these days.
Do what you like, mrclark, Google doesn't care what's actually on the page just make sure you don't bad-mouth them in your newsletters, that can get you kicked out faster than spam!
Bek.
Bek.
Back to topic. The importance of off site factors does not mean that on site factors are totally irrelevant. It is still of some importance how you use <Hx> tags.
Bek.
<H1> tag - Use it only ONCE per page?
I think that when you reach the point in your content where another H1 tag is appropriate you have also reached the point where another page is appropriate.
How is whole paragraphs of text in H1 "for the reader"?
And are your H1 paragraphs default H1 styling, or have you restyled them?
The W3C does a wonderful job of defining how something should work. But they are utterly useless when they try and tell you how to do something. And either way, why should google care when their real interest is in the content.
The average person writing or reading a website does not care about which tags are where, or if those tags meat any structuing requirement. They just want it to look nice.
If it is easier for them to use <h1> multiple times, or jump straight to <h3> without an <h2> tossed in the middle, then that is fine with them. If you want several H1 size headings, and don't want to bother with CSS, then go ahead and use H1 multiple times.
If W3C wants to waste their time dictating *where* you may use a tag, they will not change the web, the will make themselves less relevant.
A standard is not a standard because a "standards body" publishes it. It is a standard when it is truly adopted by those that use it.
W3C "depreciated" <b> in favor of <strong> for some sort of politically correct reason. Sure was a wonderful plan. Except for the fact that <b> os easier for most people to remember, and the high volume sites like Google and Yahoo have huge bandwidth bills, and those 8 characters for each use would probably add up to thousands of dollars a day for no gain.
Even better was the decision to depreciate <font> in favor of CSS. CSS sure is wonderful, but it requires learning a whole additional spec. If all you want to do is make one word red, there should be a way to do that in simple HTML. Adding CSS was a good idea, trying to force its usage instead of just defining the standard and letting people use it was a bad idea.
Trying to force "proper" H tag useage is just as bad of an idea.
Back to how Google should handle it? However they want to.
If it was me, I would probably give a page with more than one <h1> a single point on the "possible spam" scale. But there would have to be quite a few other indicators before I would ding them spamming.
Document structure is what it's all about and it's been around since before the net.
Document structure lends semantic meanin to a doc so that I, you, or a parser (Google?) can see quickly, the main topic discussed and the associated sub-topics (headers h2...h6). It enables us to get a general "feelin" about what we're gonna find, for instance, by skim-readin... which we all do.
The hx tags are like a directory tree, with your C-drive being h1 - If you orgainse your folders logically, you should be able to navigate quite easily by "followin the story" of the directory titles.
A good way to witness the semantic meanin is, if ya have Opera, to use the mobile device function on a properly structured (HTML) doc and compare to one without... you'll quickly see how important, and easy, it is to better feel what the doc is about.
I contend that hx tags do play an important role with G, though more than one will most likely be ignored... similarly, an overly long one too.
The thing is, that the tags are only tools, they are not the real structure, nor do they represent the real structure. They can be used as a hint as to what the real structure is, but that is all. By following the intended use, it can make it much easier to build your page, but it has no affect on the value of the content.
If you do not want any text of H1 size, and you do not want to mess with CSS, there is absolutely nothing wrong with starting at H2 or H3. The Hx tags are nothing more or less than <p> with different fonts. From the perspective of the Search engines, the default larger font and general use as a header is where they might gain their value, not because of any special defined usage.
Blockquote is quite useful for quoting a block of text, but if you do not want to to be presented indented like that, you can feel free to use <p> with quotes around the content.
Did you know that you can make a numbered list by just typing numbers in front of your text? For the user, there is no difference. <ol> is just a tool the same as <h1>.
The tool does not define the work. You can use a pair of vice-grips as a hammer. It may not be the most efficient tool for the job, but it works surprizingly well with a little practice.
We may hate the code that some people and programs produce, the the important thing, in the end, is the content of their message, and the presentation on the screen.
I don't hide my H1 tags with css.I have sites that people usually come to hoping to find pics and I am very proud of my original text so I make it so big that the surfers can't help but read it.
This title/header talk is indeed confusing. When webmasters talk about "titles" they don't mean the words that you read in big letters at the top of most web pages. They mean the text put in the title tags in the head section of the document. The visible "title" is mostly created in a <H1> header at the top of the body section in the document.
Headers are a nice tool to help structuring your documents. They are not really necessary, but if you know how to use them they make your life as a webmaster easier.
I use H1 for page titles and style them in my CSS files, mostly to make them smaller because I consider standard H1 to be ugly. If you agree with me but don't know CSS then you can specify the size of your H1 in the HTML file. Or you can use H2 or H3. AFAIK all these possibilities are quite allright with Google.
Then I use H2, H3 and H4 (all of them styled in CSS) for subtitles. I do not stick strictly to W3C's standards but try to make my choices so that they make sense for those individual pages where I make them.
Others have a different attitude. Webmasters have for perfectly innocent reasons many different ways of using headers. And Google know that. And some webmasters use headers to try to cheat search engines into believing that some pieces of their texts are more important than they really are. Google know that too. Make your choice. My own choice has been to use headers cautiously.
Yes. Of course this matters else lots of pr0 sites would not rank at all.
Try ranking a blank page with lots of links to it.
Ummmm .... that's exactly what is working on Google these days ....
You CAN get a blank page ranked using links but you can't get an H1'd page ranked without links (for anything relevant that is!). Which is why I'm saying don't bother thinking about Googlebot when using reader-orientated styles such as H1, bold, italics, underline etc. as the conversion rates of your customers will always outway the ranking boost by focusing on Googlebot. 1,000,000 visitors when only 10 buy is worse than 1,000 visitors and 20 buy - (you haven't said if you're site is commercial or not but I'm getting that feeling from what you're stating.)
The thread just started getting beyond my humble capacity to follow but I think BIgDave is saying google doesn't care and I can't see why they should either.
Because they're in the business of document retrieval and headers play a major role in defining what a document is about.
If you don't use hx tags, but instead use large (now deprecated) <font> tag attributes and/or <strong> tags to achieve the same visual effect, then, as the document isn't semantically structured in the eyes of a computer, a parsing algo - any parsing algo - will miss the importance and relevancy of what's being said - In the hierarchy of the document tree, the font and strong tag are amongst the lower echelons.
Consider also, accessibility: Blind users can't see headers or differences of font. They will rely on text-to-speech apps. They in turn will seek "guidance" from the author as to how to navigate the document lest the user is forced to waste time readin it all through without findin what they want... a parser is "blind" too don't forget, ie... Google etal.
Web-enabled devices are also expanding beyond the PC-based browser - Now the font tag is deprecated, your text will appear unorganised and consequently harder to read should you wish to stylise headers using this method.
By using an H1 tag, I'm tellin the user/parser that this document is all about x - A font tag doesn't have that semantic power - It's just an instruction.
If you use multiple h1 tags within one doc, you, in effect, confuse and distort the true document meaning - You're cuttin short the discussion and starting a new one midflow in essence. In SEO terms, this can only work against you; It's as if you were nicely buildin-up points (relevancy to the topic) and then you reset the counter to zero again... even though it's the same topic.
Headers act as a path to the info we want. Just study your own actions when browsin a newspaper, thumbing through a book, viewing a website or planning a long journey even. We look at it top down (topology), each step gettin more detailed until we "arrive". If your headings are not labelled correctly, or misused, you basically don't arrive or at best, "arrive" the long way.
Those that doubt the wisdom of the (w3c) intentions of this schema seem to object on the basis of (artistic) presentation and webmastery skills. Well, we're talkin about blind algos here. All algo calculations have rules and must first assume that those rules are adhered to (We can build-in error-handling but first we must establish the axiom to base that error-handling on)...
Because headers have a semantic meaning, it's entirely appropriate to look for it, and take the appropriate action based on it's meaning. It's not the job of the algo to teach the webmaster how to write a page. Nor does the algo care for how your doc looks... and what does that matter if your user views your doc through a device that's stripped away superflous tags anyway? (Try the Opera mobile thingy test).
At the end of the day, I don't know with official certainty that G makes a judgement on hx tags. My subjective research (strongly) suggests it does and it makes perfect sense that it should because appropriate headers help define what's being discussed.
If you don't use them, or use them incorrectly, then you run the risk of a myriad of operations, now and in the future, making unintentional decisions based on it. For instance, I have an app that generates a sitemap. It does this by parsing folder names, file names as well as the document title and all headings within the document. It even indents those headers on the sitemap based on the hx values so that what you end up with appears like a directory tree. This is extremely useful both for me and my users.
If you need another argument: There are, as we know, more than 8 billion documents on the web. To index all these appropriately requires huge processing power as it is. Now, based on the principles of document structure and retrieval, do you think a document with no semantic structure requires more, or less processin?
EVERYTHING matters to a degree. I personally try not to miss anything and try to take advantage of everything.
1. Common sense should indicate that you do not know the people posting. They could have 2 or 3 logins with webmasterworld. We cannot see the IP’s to know for sure.
2. Common sense should also tell you that SEO’s do not want you to know their tricks, it would give them competition.
I’m not saying if the topic of this post this is true or not, nor if anyone posting is, has, or will be a SEO, I’m just saying use caution. And if I didn’t have proof, I wouldn’t be posting this.
I never said there wasn't value in giving the headers extra attention. They are of use in helping decide what a page is about. It is just like looking for bolded words, except more so.
But you are pushing to define how someone puts their document together. And you are suggesting that if they do not know CSS, that they must follow a certain structure, even if they do not want any printing as large as the standard H1.
There are many perfectly valid document layouts. You could have an intro paragraph, then use the H1 to make the statement that is the point of your document, and close with a final paragraph.
It is not the problem of the creator of the page if your precious program is not able to figure out the meaning of the page. Browsers don't care about the meaning, so they will just render it the way it is coded.
Google doesn't care about the layout of the page either. They care about the content. They will use certain layout information to help them decide what is important, but they are not going to care if a page is laid out "correctly".
If you don't use them, or use them incorrectly, then you run the risk of a myriad of operations, now and in the future, making unintentional decisions based on it.
If you need another argument: There are, as we know, more than 8 billion documents on the web.
I think you are defeating your own argument here.
What is the breakdown of those 8 billion documents that are already out there? I'm guessing that almost none of them meet your definition of proper structure.
Like I said before, it is not a standard until it is adopted for common usage, and those 8 billion pages are not going to be updated.
And since those pages are not going to be updated, it is the software that cannot handle those pages that will be considered broken.
Even if a page has "broken" HTML, browsers will try and display it, and google will try and index it. And as long as all the major browsers do a decent job on the page, that page passes the important test.
Don't get me wrong, I am in favor of validating code to fix errors, and using good structure in documents. But you are heading down the wrong path if you want to try and define what is an "acceptable document structure".
I'm reminded of the difference between most dictionaries and the OED. Most dictionaries try to define a language, the OED tries to document the language.
You can define what an operation does, you will never succeed if you try to define exaclty how people are supposed to use it.
When people are in dreamweaver, composer, frontpage or even word, they will run through the selections that are offered to them in the menu and select the one that gives them the look that they want. Almost no one will worry about how well your program will be able to figure out what they are doing.
I have never seen anyone say something in the past to mislead on these forums for disinformation purposes. I can say that people at one time or another {including myself} have said things that may be wrong. Usually someone with the correct knowledge sees it, and they will say something to correct the post or at the very least disagree with it.
With hundreds of thousands of websites on line, I don’t think there are enough SEO’s out there to really worry about too much competition at the moment. One day that may change, but not today.
You did word your statement carefully, but I find it hard to believe that you’re correct. At the very lease I hope your not and I also hope others voice their opinion on this matter.
CygnusX1
When people are in dreamweaver, composer, frontpage or even word, they will run through the selections that are offered to them in the menu and select the one that gives them the look that they want. Almost no one will worry about how well your program will be able to figure out what they are doing.
You may well be correct, but Google to many is certainly not "almost no-one" ;) it would be very easy for a algo to sort the "wheat from the chaff" by diluting semantic, document structuring elements (like heading elements) and receive no valid argument other than "it used to work".
It also means that young tom, dick or harry who've just cranked out a site using the aforementioned technique on their collection of football cards is not going to appear #1 for a keyword because of their unintentional overuse of an particular element.. in fact if you came across a site of theirs which ranked above you would you not then be shouting the odds?
So what better way for an algo to keep everybody (except SEO's perhaps) happy by sticking with W3C, accessibility, semantics (as far as HTML can cope with them) and preferred document structure? Never mind some of the other stuff algo's can't detect, this one's easily dealt with from an algo/dilution point of view. No hidden text, just count the number on the page and dilute accordingly...
so in answer to the question, imo yes only once per page for an <h1>.. or be able to afford any dilution that may be happening.. some other elements may be more difficult (not impossible) to deduce a "dilution" factor but this one is easy.. it's usage is taught in beginners HTML, any "abuse" is either SEO or unintentional (novice), so either can be filtered/diluted with one fell swoop of an algo change..
Suzy
If you don't use hx tags, but instead use large (now deprecated) <font> tag attributes and/or <strong> tags to achieve the same visual effect, then, as the document isn't semantically structured in the eyes of a computer, a parsing algo - any parsing algo - will miss the importance and relevancy of what's being said
the original google whitepaper doesn't agree with this statement, and neither does a recent test I did on a keyword phrase, what I found was that the basic methods Google is using to classify content have not changed at all, they are still focused on text weight/size, relative to surrounding text. Out of 10 top ten sites for keyword phrase 1, 9 used either bold or font tags to make their headers, and this corresponds exactly to the way the google white paper said they classified page content. What does appear to be changing is how many pages they are keeping actively indexed, and their methods for deciding which pages to index and keep. To me it looks like they are basically doing the equivalent of cleaning out a full harddrive, set the bar slightly higher to maintain your position to weed out extra junk to trim down the bloat.
