One of the "fun" things when I worked at a big company was thinking through how to make the very first launched version of something work at millions-of-users scale. There is no organic growth in traffic from zero when a non-negligible subset of billions of active users will try it out right away.
But I think on net it was still an easier problem than scaling a startup product from zero, because the tooling and processes to support huge usage is already so mature at these big companies.
But I think on net it was still an easier problem than scaling a startup product from zero, because the tooling and processes to support huge usage is already so mature at these big companies.