Here is an important question from Micky. (Account manager at Ruby-people.com.) And the question he put on a linked in group is What are the main pitfalls to avoid for Ruby-on-Rails start-ups? After I answered him there it occured to me to put it here for everyone to review and discuss:
 
Here is my take:

1. RoR is not “replace all” solution: It is highly recommended for database driven small and medium sized Web2.0 based web products. If you want a content based website you dont need RoR, a lot simpler solutions are available. Its not fit for any Real time use too.

2. Continuous evolution: RoR is evolving quite fast and any startup need to keep up pace with the latest.

3. Only paranoids survive: Keep a strict watch on performance, scalability and security. Ruby has its own vulnerabilities and that can be (should be) taken care of.

4. Complete skill set: RoR is a part of the Web product’s big picture. It should help any startup to have MySQL, Linux (shell scripting) and UI(Flex/JS, Ajax) experts on-board as well.

5. Look for niche: For a start-up it would help to look for a niche area in developing RoR (RoR still has a huge scope of improvement) as well. Being a part of product or services clan may not help much until you really have an excellent out-of-the-world product.