First off 50% is not what I would consider a high bounce rate.

Let me get this correct. Your gonna do a new site say a "testing Site" under a new domain name work out all the bugs then upload the new design to the old domain?

SEO issues

If this is correct be sure and have in the plans for the old url's to be 301'd to the new urls.

Be sure and have the company add a robots.txt to block all bots or better yet have it password protected so the "test site" can not be indexed by any SE.

You "most likely" will experience some drop in the serps due to the overhaul but from experience the terms "most likely" will return. There is no way to tell you they will come back to the present level as with any move there is always a "risk".

I know I am using some vauge terms here but there is really no way to tell you what will happen. Some sites have bounced right back, some never missed a step, and some went into the dark hole.

Anytime you do a change in the URL structure there is always a risk factor.

design issues

This is something that will need to be tested on your end to make the site feels more appealing to your visitors and the navigation makes getting to the desired area a no brainer.

Best way to test is use a variety of friends men/women in that field. Give them instructions they did a certian search came to your site then watch them to see how easy they can move through the site without you giving "hints". Watch them watch there eye movement, make notes on problem areas, things they like or things that are a distraction. Afer the test is completed then do an interview with them see what they wwould like to see or not see and take their observations to the design team.

coding, technical etc issues

I would make contact with a outside source to check their coding "before the site goes live" to make sure they have the CMS/DB set up correctly, there are no spider traps, duplicate URL issues, security issues, and other issues that can really cause you some problems if not done correctly.

I can't stress enough to proceed with caution check and recheck to make sure the upgrade goes smoothly and wish ya great success.

You own the "old design" website. Then contract with another company to update it? Then want to sandbox it elsewhere (protect it as bwnbwn suggests!)? These things are possible, but...

Gotta ask... why not update the current website in place? Why chance losing any link juice it already has? Am I missing something? Forget SEO for the moment. What you will lose (and not much) is return visitors "startled" by the new look. If done right they will say "cool" and keep returning...and you've lost nothing.

That said, many sites that are 1996 still perform big time. Sometimes you change things, sometimes you don't. If this is a commercial site then yes... you'll have to do the research to see if SIGNIFICANT changes will up the ROI, otherwise changes can HURT.

Just a thought.

sounds good to me ....... Ive tried to seo my site meta and text ting but at some point the design needed to change,,...

if you change your layout seo will not be effected at all

guys thank you for your replies:

bwnbwn:

Your gonna do a new site say a "testing Site" under a new domain name work out all the bugs then upload the new design to the old domain?

yes!

If this is correct be sure and have in the plans for the old url's to be 301'd to the new urls.

why? hat I'm going to do is create a transition domain name that I experience what I want to have finally on it (sandbox as you said) and after all the changes I will terminate this domain name and every final design and content will be uploaded on my old domain.
the point is I don't want my clients see what I'm planning to do with current site. so I won't redirect the old domain.
the robots.txt is a must. I agree with that. I will block /

Anytime you do a change in the URL structure there is always a risk factor.

good point. there would be no changes in URL structures as far as I can provide. only the new service I'm gonna install will have new structures because it'll be completely dynamic.

I can't stress enough to proceed with caution check and recheck to make sure the upgrade goes smoothly and wish ya great success.

thank you for your supportive comments. especially the design issues part on how to realize the visitor behavior was valuable.

dear tangor:

why not update the current website in place? Why chance losing any link juice it already has?

what you miss is the fact the site is a static website right now, so I can keep it solid till each phase finishes and then I can entirely upload the new edition.
I believe my site is kind of those 1996 sites you said are doing fine, but I have to move on with new services. especially when I think of what my competitors can do (but I wonder why they don't!), I panic!

[edited by: engine at 3:37 pm (utc) on Sep. 22, 2009]
[edit reason] See WebmasterWorld TOS [/edit]

honarparvar with keeping the urls same that is a smart move and you do not need to 301 anything. I was assuming the url's were gonna change.

The only reason I said to passwword protect the testing site is sometimes mistakes happen and the testing site gets indexed before the new content is uploaded.

Case point I have a Disaster Recover Site no links pointing at it no one has a clue the site exist except me. I have a disallow meta tag a no index meta tag a no index/no follow disallow all robots.txt file, but Bing still has the site indexed. There isn't a bit of content on the front page no links just the wording "This is example.com Disaster Recover Site" Please go to "link" to the real site.

You just never know so just to be "Safe" password protect the testing site.

thanks dude, it's wise. I will consider it.