can you elaborate?

They change things. See what works. Iterate.

That's called being rational; nothing startup-specific about it.

It is being rational, but it's also absent from many other organizations. I don't think it's because the people involved wouldn't want to try something else when their initial solution is failing. It seems to be all about face saving. The culture around them is partly to blame for that. New companies have a fresh culture and the people involved know that failure is a very real possibility.

Big companies forget step three. They tend to head down a path, see what worked, then retroactively redefine goals so that everyone hits their target.

I think the startup-specific part is being willing to identify mistakes.