Also be an interesting exercise to help determine what it is you actually want to re-design and change, based on … . What questions we need to ask to better understand the need for the re-design.

For example, reviewing the site’s existing content and determining what content users need to get to faster may be a much better use of resources, than putting a new design skin on an existing problem.

I can’t give you an example of someone who did a major re-launch “well” because I don’t know what that means. Major re-launches are possible, but the risks are very high. Avoiding the major re-launch is a way to dramatically reduce the risk.

Would a site that has existed largely unchanged for so long be a good target for a re-launch, or would the users rebel from the changes all the more?

In our experience, we find users don’t mind minor changes. But we find they hate dramatic changes, no matter how frequently the site has been updated in the past.

Unless you do an amazing job of migrating them from the old design to the new, by giving them very good tools to learn the new site incrementally. However, to do that, you have to actually study how people will migrate. And the more facets of the site that change, the more expensive the research and the development will be. And the more likely you’ll screw it up.

Again, it’s all about reducing risk.

Perhaps there are web sites that are doing it so well, that its hard to see the changes Then again, if you have a solid base to start with, there may be little need for large step changes.