Business today is all about some type of Continuous Improvement (CI) process. We hear about it all the time -- Six Sigma, Lean, Agile,ITIL. Primarily based on the Demming Cycle, CI processes have been shown to significantly improve quality, and can be credited with many significant business success. (Toyota before 1999 is the poster child for CI.) The problem, however, with CI is the 'I' -- improvement.
When you improve something you you take something that exists, and you make it better, but you have to respect the law of diminishing returns. For example, if you have 5% marketshare, it's possible to double your marketshare over some amount of time, but if you have 80% marketshare its impossible to double it. (While I used marketshare as an example here, there are physical limits to all processes that apply in the same way.) In other words, at some point you won't be able to significantly improve a process/business/product, and in order to grow you are going to have to change.
Implementing a CI process is a change, but having one does not absolve you having to change again. Frequently I see business with CI processes that begin to believe their own propaganda about quality, and improvement, and think that the CI is the change. (Toyota of the 2000s is an example of this). Small companies suffer from this as well, typically as a result of growth. The old systems, organization structures, and process that worked once, no matter how they are improved, cannot meet the new realities. But, in order to preserve 'the culture of CI' that is currently in place nothing is really changed.
Change is no fun, but if you don't want to end up on the decline end of the business life cycle you have to do it. But remember change, that will extend you growth curve, is big, risky, destructive, and somewhat painful. While I don't suggest you do changes like this regularly, you do have to monitor your business, and when you see the signs of diminishing returns you need to act.