[oarsinsync]

Sure, meanwhile in part 4 he makes another leap making assumptions about server intention:

> Our sense was that our weak team members looked at their tips — which of course were close to normal since most people don’t adjust their tips much for bad service — and concluded that, in spite of what we were suggesting, they were already doing sufficiently good work. If I’m good enough to make my tips, it’s obvious that I don’t need to improve.

Which from the business perspective, is a real problem.

Or to break it down without the assumption of motivation

> Because tipping correlates weakly to service quality, and because individual tips are always subject to interpretation, tipping removes the incentive for poor performing servers to improve.

EDIT: or read part 5 if you want some explanation behind the leap in intention. It helps to read the complete articles! :)