20 replies to this topic

[ranch]

See my questions below!

Many people seem to believe that using frames is a big no-no if you want your site properly indexed. Read below to learn why this is not true and how to achieve high rankings for a frames site.

We probably all have visited or even created framed sites with an index-page containing the famous sentence "You browser doesn't support frames ...". Many frame-sites don't even have a NOFRAMES-tag containing links to the inner pages and KW rich text explaining what the site is all about.

Of course it is very difficult for SEs to index such sites properly, but this should not lead to the common misconception that a framed site won't rank well.

Using the NOFRAMES tag will help, but problems with bookmarking framed pages remain as the url of a framed site is fixed to that of the page containing the FRAMESET definition.

On the other hand framed sites have some advantages in the design area over regular non-framed sites, i.e. you can have part of your page fixed (menu etc.) whereas only the content is scrolling. Also a frames-page need not contain html code for navigation etc. which would probably make it more SEO friendly as the text/html ratio is higher compared to a non-frames page.

My own site www.fejo.dk (not in English) uses frames and I get lots of traffic due to good rankings on my KWs. The site has disadvantages though. The url is fixed, giving problems with bookmarking and even worse: When you find one of my inner pages on a SE the page is seen without the frameset, i.e. no heading and navigation, just the text. The page is orphaned.

There are several solutions to this orphan page problem where a page is shown without its frameset. Most of them don't really solve the problem. Thanks to Jill's wonderfull newsletter I stumbled over a technique called Self-Referencing Framesets explained by SearchMechanics. As far as I can see this technique solves ALL problems with sites using frames, leaving only the design advantages which my site relies on.

I then decided to redesign my site technically using Self-Referencing Framesets and while I was at it improving SEO and overall design. My new site fejo.pentia.dk (in english) is still under construction, but the frames part is working. Because of the many languages I needed to reload all framed for every new page requested. For a single-language site this is not the case, only the content-frame is loaded when a new page is requested.

The essence is though, that Self-Referencing Framesets are working great. My site behaves just like a no-frames site (every page can stand on it's own) and I enjoy being able to keep part of the design fixed for my maps and slideshows. The solution is a tiny javascript that you can implement in hours.

As for the SEO part: Now that every page is responsible for drawing the frameset itself you need no index-page containing the frameset meaning you can forget about the NOFRAMES tag and just do a normal SEO job for every page. This is the forum to learn how.

The interesting part though - are there any SEO disadvantages I've missed?

QUESTIONS:

When using Self-Referencing Framesets correctly will the spiders be able to tell at all that the site uses frames?

Is it true that a frames-page is more "SEO friendly" than on a no-frames page as the text/html ratio is higher on the frames-page because it need not contain html for navigation, header etc.?

If a frames-site is build correctly are there any SEO-downsides, i.e. will the Seachengines rank a frames-site lower just because of the frames?

Henrik Ranch

[Scottie]

Hi Henrik-

My opinon on frames is that they are only useful for online applications and extremely content-heavy sites (think government or research documents) so that you can maintain "controls" in a consistent place on the page.

There are usability issues with frames- typically there is a footer on the page. Depending on how your content "breaks" near the end of the page, people may not realize there is more to see unless half the words are cut off by the bottom frame.

Toss in to that scenario that designers love to change the scrollbar color to "blend" in to the screen, and suddenly your valuable content has a lot less opportunity to be seen and read.

From an "ease-of-coding" perspective, using SSI will give you the same benefits as frames and are even more flexibility. You can still have 1 file to change for your primary navigation and also set up numerous submenus and other "reusable" chunks of code and call them when you need them.

[Alan Perkins]

Hi Henrik

Since I wrote that article I owe it to you to answer your questions.

When using Self-Referencing Framesets correctly will the spiders be able to tell at all that the site uses frames?

Yes. The insertion of JavaScript between the <head> and <body> tags is a big clue. Also, in theory they could actually read your JavaScript and see the frameset in there.

Is it true that a frames-page is more "SEO friendly" than on a no-frames page as the text/html ratio is higher on the frames-page because it need not contain html for navigation, header etc.?

No, IMO. Because a normal content frame contains little/no navigation, it provides no clues to the site structure, preventing search engines from doing their job.

By contast the Self-Referencing Frameset does provide navigation in the content frame (it's in the <noscript> tag) so it gets round this. But I prefer to view the Self-referencing frameset as a means of creating a frameset page from a noframes page - not the other way around.

If a frames-site is build correctly are there any SEO-downsides, i.e. will the Seachengines rank a frames-site lower just because of the frames?

Possibly. In the long term it's impossible to say.

The self-referencing frameset article was written for people who absolutely have to use framesets. I still wouldn't recommend them.

[ranch]

Hi Scottie & Alan

Thank you very much for your answers. Very interesting. I understand it's very difficult these days to find somebody in favour of using frames on a regular site.

To Scottie: It's true that frames requires you to think carefully about design, so that no one (in the target audience) is left with nonscrollable frames where they don't see the bottom/right side. But IMO that's merely a design issue. This also goes for the scrollbar problem you mention. All in all not reasons not to use frames but reasons to design carefully. And that's always important!

As for SSI I'm sure you're right. I'm not very technically minded (I just know what the abbreviation stands for). I'll look into it, so thanks for the tip.

To Alan: So you are the person responsible for my site still using frames! Actually I was about to leave frames when I read your article. And I am very thankful to that article because I like the design of my new site - until someone convinces me that frames, for hard SEO reasons are a no-no. Allow me to comment on your answers:

OK, so in theory spiders could (obviously) tell that a site uses frames, but that means they don't care now I assume, and why should they. Combining this with your last answer that there might possibly be SEO-downsides to using frames (none listed) I assume that when pro's, knowing a lot more about about the topic than I do, can't come up with "hard evidence" why not to use frames, I guess they pose no problem as long as you do thing right.

I'm not trying to make you say "we love frames", but could we, today, agree that using frames or not is primarily a matter of taste, requires a (more) careful design and insight on how to solve the SEO and other technical problems (e.g. url) that frames create?

[Matt B]

I'm not trying to make you say "we love frames", but could we, today, agree that using frames or not is primarily a matter of taste, requires a (more) careful design and insight on how to solve the SEO and other technical problems (e.g. url) that frames create?

I might agree with that, but I have a like/hate relationship with frames - no love at all. At one time I was a frames evangelist, but i've recently found myself "off the wagon."

Working at a web shop where frames were the norm, (NORM!, heh, ok sorry - old reference) we are now experimenting with many non-frames solutions. The last couple of sites produced have been non-framed static/dynamic sites. We are still evaluating the differences, if any.

One of our main internal observations is the amount of time required to optimize framed sites - it may or may not be very economical for the client or the agency, depending on your view.

IMO, noframes text is going the way of meta keywords, SE's don't seem to mind if the site is framed or not, and Alan is ready to throw himself off a cliff because his article made you stay with a framed site. . .

[Ron Carnell]

To frame or not to frame?

Ask a random sampling of designers that question, and I suppose you'll get a spectrum of answers. From the viewpoint of some, there might well be valid reasons why frames are a "good thing."

Ask that same question to a sampling of visitors, however, and I would be very surprised if any of them advocated the use of frames. I can't think of one direct benefit for the visitor, though I can certainly think of several perceived disadvantages to them.

So, maybe the real question should be "Who do you most want to please?"

[claybutler]

Just implimented the self referencing frameset on my site. It seams like a superior choice to the "refresh to index" javascript I have been using to prevent orphaned pages. I've put up a test version ( Claytowne Test)and I can't figure out why the navigation reloads when you click one of the navigation menu items but all the links in the body frame pages just refresh the body frame and leave the navigation as is. It also does a one-time "open in new browser thing" the first time you click a nav link then continues to work fine after that ( in the new window). Any advice would be helpfull.
Thanks,Clay

[Jill]

Welcome, claybutler!

Sorry, I can't help you with your frameset problem, but I'm sure somebody here will chime in. Alan is in the UK where it's pretty late, but maybe he'll see this and post in the morning.

Jill

[ranch]

Hi Clay

I've noticed that you put the writefs.js include in the HEAD section. According to Alans article it belongs between the HEAD and BODY section. Maybe this helps?

Henrik Ranch

[claybutler]

Thanks, but it didn't have any effect ( putting the code in the head AND body) Claytowne Test PageI tried every combo I could think of and the best I could do is make the site not work at all.

I still get an "open in new browser effect" when you first click on a navigation button and the nav buttons continue to trigger a nav reload every time after in the new browser.

Since the links in the bodyframe all work as intended, I suspect the problem lies within the navigation page or the writejs page.

Thanks,
Clay

[Denyse]

Sorry but whats SSI ???

Denyse

[claybutler]

Server Side Includes

[Denyse]

Oh ya I've heard of that but still dont know how to use it

Any idea where I might get some info re SSI

Denyse

[claybutler]

SSI Tutorial
This looks like a good place to start.

[Denyse]

Thanks Claybutler,

Guess what I'll be studying this weekend!!!

Denyse