[unalone]

No, it's that most people don't want to use a product that changes often, if at all. They get it for what it is and if it changes, it means disrupting their way of using it. When you're working with a smaller product, you can iterate more and work closely with the people using it. When it gets larger, any change you make will alienate some users, so you have to plan to minimize that alienation.

You can change and roll back features quickly, but that gives an impression of instability among users. If things are constantly changing, they'll seek out something more stable. It's "lowest common denominator" thinking. It results in something that nobody dislikes, and that's the goal of larger groups. Niche companies are able to work much better, but even then there's some slowdown.

[jumper]

I think this is particularly important given that often the biggest cost in using a new product\tool is learning to use it well... if it's changing all the time, with no release schedule or no attempt to stabilize the main release via betas\QA\etc, then it gets very frustrating to learn.