Hello everyone,

This is my first visit / post here (please be kind!) I'm a novice developer and am in great need of advice on a site I'm developing for friends.

I have it in frames with a splash page entry. I know folks either love frames or hate them, and the splash page is strictly for the search engines' benefit. My problem is, the owners seem pleased with the design; but, are concerned about it's not showing up in the search engines. I've put everything I can in the META tags without "stuffing" them, and on every single page of the site.

I stumbled on TECHpatterns, while researching frames and search engines. I think links to sites are disllowed here; but, what is the general consensus in developing? Frames - yes/ Frames-no?

Any suggestions appreciated.

ada
Back to top

Well, we'll be merciful.

You have basically done it all wrong, every step of the way.

So, stop using meta tags for seo, first off. Use them for what they are designed for, descriptions. keywords I usually don't even bother with any more.

It's been years since any real search engine used the information in meta tags to position your site in the results. Google uses the stuff in your description meta tag to display a blurb on their results page if nothing else on the page presents itself as a logical context. But they do not use the meta description for any other purpose, especially not serp placement.

So forget that idea, clean up those spammy meta tags, make them read like a real description, and forget about the keywords altogether.

Also along these lines, don't bother trying to put key words in other places, like image tags, etc. Or in noscript tags. These are all old tricks, that will usually just end up getting the site penalized long term.

Forget frames totally, now. Dump them. Get rid of them. It is only through exceptionally careful programming that you can get search engines into the target pages, and searchers into them and return from them. So just forget that idea completely.

Frames suck for search engines, period. Forget it. They can be made to work, but as your site develops, you'll end up dumping them once you understand how limiting they are, so just skip that step and forget frames.

Splash pages really really really really really suck.

You're completely wrong about search engines liking them, they hate them. Nobody will link to a splash page deliberately, so that's a totally worthless item. Users dont' like them, because it's just one more obstacle to get to the site's content.

So why do they still exist? Amateur web developers, designers more interested in splash and flash than functionality and creating a good user experience for the site visitors. Remember: splash never makes the user experience better, it's a totally pointless display, and is simply annoying after you've seen it once, and usually it's annoying the first time too.

So basically you did it all wrong, the client's site doesn't rank. That's not surprising at all.

Also, after this stuff is fixed, keep in mind that google has a waiting period to rank new sites/domain names. This is known as the google sandbox. Google the term to learn more, there's lots on it.

But don't bother trying fix something that is this broken, it needs to be redone.

Remember, the reasons you had for using frames are bad reasons, we've probably all had those bad reasons, I know I did.

Most common reason? Inability to create a simple menu include file that is delivered through some web programming language like php or asp.

Second most common? A completely mistaken idea that a web page should scroll within a scrolling window, and that the whole site page should always be contained in the browser window, in a frame. This idea is simply wrong, but it's one that designers have a very hard time letting go of.

Remember: a single web page works best, a frame is at best 3 pages, sometimes more. Frames perform radically worse than a single web page in tests I did on that, not even close for end users.

Pages should scroll down, and if they are too long, you should break them into separate pages.

After this is all fixed, you can start worrying about actual seo issues, like creating high quality, inbound, one way, backlinks, from reputable sites. That's the result of having unique and reasonably valuable content on the site, of course. That automatically excludes 99+% of all the websites on the web.

Real seo is creating real content that shows google over time that your site is valuable, at least a niche hub/authority type site.

With clean html/css code, and all that fun stuff. With relevant weighting of key words in the title, body, and header text.

Remember, if a site is generic, a clone of another site, uses duplicate content from other sites, forget it, you'll never rank.
Back to top

Hi, and thanks for your 'merciful' candor :).

So, what do we think of iframes.

ada
Back to top

About the same as frames, except iframes suffer from the extra problem of not being readily sizeable to the visitor's browser window dimensions.

Since an iframe functions basically like an image tag, getting it to size relative to the window height requires some css tricks that aren't all that reliable cross browser.

And even with that, it's exactly the same problem for seo, you have a container page, say with the header and navigation, then iframe content, so you're right back to where you were with frames, how to get the visitor to the page that's listed in the search, how to reliably get the search engine to the page, and how to let the visitor get back from the page to their search.

You can, like me, try to resist this problem for as long as you possibly can, or you can just skip this step and stop using iframes or frames.

If you want scrolling stuff, just use divs with overflow:auto, although those are not very user friendly, and will in my opinion soon be abandonned commercially simply because average users will get too confused by them.

The simpler you make it, the better it will work. This also applies to onpage layout code versus onpage content. The more content the better. With proper CSS styling, you can often get a standard page down to about 5-8 kiloBytes, including content. With less layout code, the search engine finds what it is looking for quite easily.

However, this is not nearly as important as fixing the frames issue.

Another thing to avoid is getting all the page urls indexed, then changing them without using proper URL rewrites to the new urls. Ideally you should never change urls, but if you have to, you must add a 301 url rewrite, which is fairly easy with Apache mod_rewrite, but pretty much impossible with shared IIS hosting.

Which is why there's another, generally unspoken, seo rule: don't host on windows IIS servers, host on Apache.
Back to top

AAAAARRRRRGGHHHH! Somebody remind why, oh why do I want to learn this stuff?

So, basically - in order to make things work - I have to give up everything I thought I'd love about designing; or, become an expert programer....?? :(( not fair!

Well, I really do appreciate your time in responding; and, I've begun (yesterday) to redesign the site in question. I'm just using plain ol' scrolling <gack!> pages.

ada
Back to top

All you have to do is learn this simple formula: the web is not print.

Repeat that several times a day until it sinks in.

Then realize that what a web page actually IS is a bunch of code, of varying complexity, that essentially tells the user agent, aka browser, search spider, etc, what to do in terms of the data in the file.

You'll find, once you stop approaching the web as print, which is what almost all designers do before they figure out that no, in fact, the web is not print, that designing for the web is easy.

To understand what the web actually is, simply look at a few things: your web page, the search engine spider, the search engine algorythm, and your browser.

Your web page is not a graphical object until it hits the browser the client is using. And even then it doesn't have to be, to see what search engines see, use a text mode browser like lynx and you'll soon wake up and understand what this is all about.

A web page is a 100% linear series of commands that tell the browser what to display in what order etc. A search spider requests this html file, which is pure text, ASCII in this case, and processes it. It assigns certain items higher values than others when deciding what to do with the data in the page.

For ranking purposes, it more or less ignores almost all the html on the page, except for stuff in <title>, <h1-6>, <b>, and some other stuff. Links are given extra weight as well.

And that's it, once your html has largely been stripped out, the search spider is looking at a linear sequence of text data, with some items given more weight than others based on increasingly complex algorythms, that for most amateur web designers they are much better off not trying to figure out or manipulate since doing it wrong can kill the page and then the site.

The gui stuff, the pretty layout, that's absolutey meaningless to both searchers and users, despite what designers like to think. The overemphasis on design is something that I'm very happy to see is proving to have been a late 90's phenomenon, and is fast fading as more and more serious websites learn that what users want is information.

There are exceptions to this, like myspace for example, but that caters to children and teenagers, so it's not a particularly relevant example, but there too you'll see scrolling pages. In fact, you'll have to surf a long long time before you see a major site that uses either iframes or frames at this point. And there's a very good reason for that.

So just forget the whole page sitting wrapped in a frame in a browser idea completely, it's based on a totally false understanding of what a webpage is, the browser is already the wrapper, that's the core of the error when you think about this. I made exactly the same mistake by the way, for years.

Let the pages scroll, you will then be creating what users and search engines both want. And then realize that the designer, and sometimes misinformed clients, are actually the only people to want what is not good for their users or search engines.


LOL, no, you don't need to become expert, but you do need to learn how the web and web pages work.

Adding includes to a page is about 1 line of code, you can find it anywhere, and adding php processing to apache without changing the url of the pages to php and setting an include path is about 2 or 3 lines of code in .htaccess.

And that's all that you have to learn.

But really, the ideal solution for designers is to find a reasonably competent programmer to work with, since the web is becoming increasingly programmed. The hand made website is becoming more and more rare, most sites now are run by either forum, blog, cms, or some combination, software. And the only role for the designer for those is creating custom css templates for the layout. That's a one time thing, and requires knowing how to deal with templates and css. Especially css.

There is no way around this, more and more the web is automated, and that's how it's going to be, so if you want to have and work for clients, you'll have to adapt to this reality long term, even short term.

No client benefits by having no programming on the site except for very rare cases where they will be the only ones ever updating the sites, using dreamweaver or some other equally destructive wysiwyg web page software. I haven't made a non programmed site for years, even if it's just having includes for shared elements like navigation bars.
Back to top

Here's a sample of what this page looks like for example to a spider, more or less anyway:



and that's about it, not very interesting.
Back to top