Pretty basic question here...
| 5:34 am on Sep 11, 2000 (gmt 0)|
I'm toying with the notion of using frames to promote client sites in the engines...reasoning being that I'm going to host the "promotion" sites on my servers and redirect through a tracking code to their actual site(s). I don't want to use cloaking, nor do I wish to use the deadly meta-refresh or java refresh scripts. So I figure I can use a blank (hidden) frame and have the clients home page load in the main frame, thus taking up the whole browser window.
Questions - I know that placing the appropriate meta tags, content, etc. in the <no frames> section of the pages' HTML is the way to go here, but I'm not sure which engines will respond in which manner to this technique. Danny Sullivan's site notes that some engines support frames and some do not...will those that do not support frames read the "no frames" portion of the page, or will those that do not support them read it? Or will both read it? Are there any specifics on various engines and their treatment of framed pages? Finally, is there anything I'm overlooking here? I've hated to use frames in the past, and honestly I'm not even very familiar with how they work beyond the basics...any help is very much appreciated.
| 3:20 pm on Sep 11, 2000 (gmt 0)|
Between the engines that support frames and the engines that will read the NOFRAMES tag, you will get noticed. How well is the question.
About six months ago, I was looking at the inital rankings for a new site where I had "good reasons" to use frames. So I had used a system of:
2) NOFRAMES content in the frameset page to mirror the content page
Whoops! I soon discovered the pain of always remembering to update the NOFRAMES section everytime I made a small edit to the content page.
Nevertheless, I was relatively happy with most of the rankings. Relatively. Google was top 30 on the main keywords, Inktomi (well-behaved at the time) was top twenty, Alta was decent - at least some in the top 3 pages.
The customer was not getting enough traffic to break a profit. I struggled to improve the optimization of what I had, and got increasingly tangled in the interconnections of all my code.
Finally, I had one simple clarifying thought: why not bite the bullet and redo the site as flat pages? Give up that "nice" navigational interface for the sake of getting more traffic?
Within three weeks of submitting the new, unframed site, search engine referals had tripled. Pages that had been in position 22 on Google were now in position 5 or 6. The content was the same, just the frames were gone. Other engines showed similar boosts. Inktomi began having digital spasms about that time, so I can't really say what the effect was there -- newly listed pages just got buried for everybody.
Your mileage may vary, of course. You definitely can get decent rankings for a frame site if you are willing to do a lot of work, and not onetime setup work but ongoing work -- updating the NOFRAMES section every time you make an edit is a big pain in the nether parts.
My overall feeling is that the exact same page content will still rank higher if it is not in frames of any kind, even invisible ones. Not only that, but your site maintenance will take half the time.
If there's any way to rethink your strategy and not use frames, I suggest you go with that.
| 1:07 am on Sep 12, 2000 (gmt 0)|
I had the same feeling. I've gotten some results with frames pages, but even better results with non-frames. I'm pretty new at this, but have done pretty well for clients thus far. Now I really want to step it up a notch.
If anyone can tell me what they think the best way of solving this problem is, I'd really appreciate it: essentially, I will be hosting sites on my server that will be optimized and submitted to the engines. These sites must direct traffic to the clients' real sites at some point or be direct copies of their sites (with their permission). Ultimately the hits from my optimized pages must be tracked because clients will be paying per search engine click. So what do you all think is best? To reiterate, these "sites" are not meant to be actual sites, but just traffic generating domains for clients' "real" sites. I'm not a designer, but I can make a fair page in a WYSIWYG editor.
1) simply use replicas of their sites with all the functionality of theirs, so traffic never actually has to go to their real sites, but they still see the benefits. I'm not partial to this one and will probably not use it.
2) cloaking. This is only slightly better than #1 above, simply because it requires constant monitoring of the the spiders visit pages, new agents, etc.
3) meta refresh...that's just bad news in my opinion.
4) java refresh...same as meta refresh I think
5) have a "click here to enter" link on the gateways - again, disgusting option that looks unprofessional and may turn off visitors, thus losing clicks
5) frames...again, causes a problems it seems.
Does anyone know how to effectively forward traffic without pissing off the powers that be in the search engine arena and withoug sacrificing quality of optimized pages? It really seems that optimizing the clients' actual sites or using cloaking are the only two options - am I wrong here (please tell me I am)..
| 3:26 am on Sep 12, 2000 (gmt 0)|
I can't tell from your details exactly what you are setting out to do ... is this some sort of doorway site for each of your clients? There are a lot of issues relative to keeping a theme relevance that could matter a lot, unless your clients are all in the same field.
At any rate, one thing to be wary of is duplicating pages that will also appear under another domain. Some of the engines will be very unforgiving if they detect duplicate content.
When I wrote about my recent frames experience, I actually shrotened the story a little bit. The original site was multiple frames ... a header with primary navigation, a sidebar with secondary navigation and the page content. As a first step, I simplified the structure to a basic two-frame set up. That didn't help anything at all. The jump in rank came only when I completely got away from frames altogether.
| 4:43 pm on Sep 12, 2000 (gmt 0)|
Basically what I'll be doing is setting up a new domain and IP address for each client. Then a site will be designed for each client that includes relevant content, a strong theme, and optimized pages. The content will not be an exact duplicate of the clients' real sites, but will include the same overall theme.
But each "click-thru" that comes from the engines will have to be tracked and the traffic will have to ultimately land on their "real" sites (not the site on my server). So I have to forward the pages I'm creating on my server to their site without having an ugly manual re-direct, getting the domain/IP banned by the engines, or sacrificing the optimization of the page.
So I'm not worried about engines picking up duplicate content on multiple servers/domains, or about themes, since each client will have a "themed" site with a unique domain and IP address. What I do worry about, however, is what I referenced above - how to move that traffic along to their "real" sites without being banned, without sacrificing optimization, and without the manual (click here to enter) re-direct.
Thanks in advance...
| 7:46 pm on Sep 12, 2000 (gmt 0)|
>>each "click-thru" that comes from the engines will have to be tracked and the traffic will have to ultimately land on their "real" sites (not the site on my server). So I have to forward the pages I'm creating on my server to their site without having an ugly manual re-direct, getting the domain/IP banned by the engines, or sacrificing the optimization of the page.<<
I think I get it. However many pages are in your version of a client's site, the goal is not to get traffic to read your pages, but to read the client's pages. Right?
If so, I can see why you were thinking about frames -- the "invisible" 0% frame would come from your server and the visible content would come from your client's server, right.
I can see how this MIGHT work, but the pitfalls are many, and as you've mentioned, getting banned is no joke. Anytime you try to feed a spider something different from a human viewer, you are running a risk.
Send me an email if you would like to discuss further (see my profile)