pjh
"A crude instrument like this is a pretty good start. It begins to shake things up and overcome the internal inertia." http://www.mckinsey.com/insights/strategy/breaking_down_the_...
There will always be an epsilon to conquer; most of them aren't worth the effort after the first couple approximations, and I value accuracy over precision.
"It's better to be vaguely right than exactly wrong" — Carveth Read (1920, predates Keynesian equivalent, q.v. https://en.wikipedia.org/wiki/Carveth_Read )
We support our 4,500 active users with a combination of a shared gmail support account, admin screens on the core web app, and a tiny globally distributed team.
One day that won't be sufficient, but it's seen us through the first five years.
Our original intention was to automate some of the standard interactions, that we would just manually process the transactional emails 'for a while'. But funny thing is, the human interaction has proved to be a critical and valued part of our service. We’re pretty sure those actual humans will always be part of the service Beeminder provides.
[1] http://tagti.me/