52 replies to this topic

[Ron Carnell]

Where ever did you get the idea that I was replacing <p> with <div>?

LOL. Apparently I need to work on my communication skills.

There are no layers on my page,
Nor windows, curtains, locks or doors.
Green eggs and ham can best explain
Our common use of metaphors.

The tools I use each have a name
and answer not to eggs and ham.
Yet, still they work out just the same
for those who just don't give a ... darn.

[Mel]

whatever

[websage]

Going back to the question that started the discussion...

Using tables for layout often results in using complex nested tables which tends to bloat the code. Separating content from presentation and layout (when using <div>, or layers as the Dreamweaver users call them, for positioning) tends to simplify the page's code structure. As a result, the meaningful content tends to go higher in the page body. This should increase the chances of the content being indexed.

The same benefit is often achieved when extracting a JavaScript into an external file -- less clutter results in easier to index page, thus in higher rankings.

Having said that, yes, you do need to know your customer and decide if you want to do all your layout in CSS.

At any rate, even if you all you did was remove the font and color definitions from your HTML code and put them in their proper place, a CSS file, the resulting page would be with smaller file size, thus faster to download, thus easier on the impatient web crawlers when they attempt to scan your page.

My 2 cents,

Mitko

[Ron Carnell]

As a result, the meaningful content tends to go higher in the page body. This should increase the chances of the content being indexed.

Should be easy to prove. Find a page less than Google's maximum that is only partially indexed. That would clearly indicate that content placed high in the document has a better chance of being indexed. My experience has been, however, that if Google indexes a page at all, it indexes the whole page, not just the content nearest the top.

There are lots of really good reasons to use CSS, and some of those reasons can even affect ranking. I see no reason to believe, however, that lighter use of HTML tags or getting the content closer to the top of the file (not the same as the page) will have any impact on SEO efforts. From the standpoint of programmatically parsing a page, it just doesn't make sense. Maybe I'm wrong, but I really think this is just another of many [url=http://www.highrankings.com/forum/index.php?showtopic=15499]Seo Myths[/url] with no real evidence to support it.

As much as I love tinkering with code and design, I'd rather spend my time building content. I know the effect that has on rank.

[qwerty]

Think of how many people use that trick of inserting a dummy table cell in order to put content above navigation buttons. The same intention is behind the practice of placing content into a layer at the top of the body.

[Ron Carnell]

Et tu, Brutus?

CSS2 can be used to similar effect, Bob, but I don't think that's what we're discussing. What you're describing is an old trick for rearranging content, putting one thing that is going to be indexed in front of something else that is going to be indexed, in large part to increase its importance. That's a perfectly valid technique, which is why I differentiated between top of file and top of page in my last post. It's not, however, the same thing as trying to minimize HTML code in an attempt to get your content nearer the top.

Again, don't get me wrong. When I started programming, 8K of RAM was a good sized machine and I'm just as obsessive about saving a byte today as I was then. Frequently loaded graphics on my sites typically have a two-character name and reside in a three-character directory, just so I can save the few bytes associated with longer names. With millions of page views, those few bytes matter. I fully agree we should use all the CSS at our command to make our files lighter. But NOT because it will increase SE rankings.

[qwerty]

Sorry, but could you clarify that? As I see it, content at the top of the file gets indexed first, content at the top of the page gets read (by humans) first. What I like to do is put that content into a layer so that it gets indexed, loaded and read first, no matter what the visual look of the page may be. So it loads before logos, graphics, navigation, etc.

Are we not talking about the same thing?

[Jill]

There are lots of really good reasons to use CSS, and some of those reasons can even affect ranking. I see no reason to believe, however, that lighter use of HTML tags or getting the content closer to the top of the file (not the same as the page) will have any impact on SEO efforts. From the standpoint of programmatically parsing a page, it just doesn't make sense. Maybe I'm wrong, but I really think this is just another of many Seo Myths with no real evidence to support it.

Here, here! I agree completely.

The search engines aren't dumb (well not that dumb). They know how to ignore code that is not relevant to their mission, and they know how to read a whole page as necessary. Since when in life is only the stuff at the top of the page important?

Since never.

Jill

[Ron Carnell]

Sorry, but could you clarify that? As I see it, content at the top of the file gets indexed first, content at the top of the page gets read (by humans) first. What I like to do is put that content into a layer so that it gets indexed, loaded and read first, no matter what the visual look of the page may be. So it loads before logos, graphics, navigation, etc.

And that makes perfect sense, for what I think are actually quite a few good reasons.

Some seem to feel that using CSS2 to replace tables would result in less HTML code. I don't think it would, but let's assume they're right. Would less HTML code then get the content better ranked just because the content was closer to the beginning of the file?

Let's take an extreme case that eliminates some of the techno-babble of CSS2 and positioning. Image that your external CSS code and maybe 200 lines of external JavaScript were inserted into your <HEAD> tags as was done five years ago. That pushes your content WAY DOWN into the file. Now we would never ever want to actually do that, because it would certainly affect the user experience, but the questions remains -- would having the content buried so deeply in the HTML file adversely affect SE rankings?

A lot of people seem to think it would, but I've just never seen any evidence to support that idea. The first thing a spider is going to do is completely throw away all that extra code, just as if it never existed. Takes one line of code to make it go poof! The order of your content matters. Where your content sits in the HTML file doesn't.

Jill, thank you, thank you, thank you. I am obviously not communicating my points clearly enough, and that's my fault, but it sure feels good to be reassured what I've said isn't entirely in Greek.

[qwerty]

I think I get your point, Ron. And if I do, I think we actually agree with each other.

Of the code a spider actually pays attention to, we agree that what comes first matters, right?

[Jill]

Of the code a spider actually pays attention to, we agree that what comes first matters, right?

I don't.

[Ron Carnell]

Of the code a spider actually pays attention to, we agree that what comes first matters, right?

Yes, we agree. But, uh, with qualifications?

I actually think that what comes first matters more for the visitor than for indexing purposes. I don't want a visitor sitting there watching my navigation download. I want them reading content while the navigation loads.

As to search engines, the first Real Content in the file has the best chance, across the SE board, of being the text that is displayed in the SERPs. That makes it about the most important advertising copy you'll ever write. I believe a #7 position in the SERP with good copy that entices the clicker will draw more traffic than a #1 with lousy copy. Used to be we could absolutely control that ad copy. Now, we just put our good stuff at the front and hope for the best. But that still justification to put it at the front, I think.

Rankings? I don't think I'm as quick to dismiss the possibility as it appears that Jill is. I know that some believe the first sentence or paragraph the spider finds is potentially more important than content found farther down the page, and I can certainly see the search engine's logic if that is true. Go to an English professor and he'll tell you that text is called the topic sentence. I can tell you with almost certainty (as certain as anything is in this business) that AltaVista once gave added importance to keywords found in the footer links of a page. So, yea, I see the possibility that the position of content can affect rankings.

I really don't think it could be weighed very heavily, however. I consistently manage to rank sites well with lead-in paragraphs designed not for SEO, but to hook the reader into continuing down the page. Could I possibly do even better if I used a topic sentence like we learned in college? Maybe. The things that aren't weighed very heavily in the algorithms are nonetheless accumulative, and one or two more points to my score rarely hurts. But most of those points also carry a cost, and each of us has to decide how much we want to pay.

[dimok]

Wow! So rich discussion.

BTW, do you have any example when pages without tables have reasonobly high ranks? If you so defending CSS2 usage let's do an experiment:

make two pages with identical content, but one with tables layout and another with CSS2. Submit both to Google and see, which will be higher?


I think that what we are talking about have a very little influence on ranking. So, don't worry

[websage]

Just another article to add to this discussion:

Using XHTML/CSS for an effective CSS campaign

See if you agree/disagree

Mitko

[rabbit]

Quick note, this is my first post here, so I apologize for coming off somewhat harsh, but I have to say this.

WOW! I signed up for this board specifically to talk to Mel -

You're nuts! HTML DIV tags are nothing but HTML tags without CSS. A "layer" is born only through the use of CSS-P (CSS Positioning (which in turn is just an absolutely positioned CSS element (be that a DIV, P or even a TABLE))).

And! To the person who had said that page layouts cannot be done in CSS without layers, needs to visit Glish...

http://www.glish.com/css/

In fact, I think everyone in this thread (save an elite few) need to go there and educate themselves on the power of CSS.

For those concerned with out of date browsers, I have two things to say:

1. Target audience matters (I know someone else pointed this out, and kudos to them)

2. There's a strong web effort being made by WaSP to encourage people with out-of-date browsers to upgrade. (more info here: http://www.webstanda...t/campaign/buc/ )

Whew! I feel better.

Now that that's over with, I have to say I like this board, and the discussion made here. Hopefully I won't develop a following stigma by making this post.

Anyway, a little bit about me... I was hired by a relatively large company as a marketer, so I'm learning as much as I can about how I can "market" the various webpages this company operates. However, I'm a developer at heart. I program in PHP/MySQL, I do graphics, etc... Also, in case you haven't noticed, I'm a strong supporter of CSS (as well as other newer technologies).

So if you ever need help with anything concerning CSS, I'd be more than happy to help you out as best I can!

- Rabbit

PS:

I know my rant has little to do with the actual topic discussion and I apologize.

Edited by rabbit, 22 August 2003 - 05:26 PM.