Unless we have good reason to believe the standard CEO-as-monarch structure is perfect, experimentation is a good thing. Moreover, his basis for suspecting his

Unless we have good reason to believe the standard CEO-as-monarch structure is perfect, experimentation is a good thing. Moreover, his basis for suspecting his particular experiment might work was reasonable, since he was basically applying by analogy a standard neoclassical model of how a market at large works to his individual business. Before he implemented the structure, would I have expected it to fail? Yes, but I expect most start-up ideas to fail, too. You never know. We can't forget that we learn as well from negative results as from positive results. Solid analysis of Lampert's failure here helps all of us, because we can see a little bit better what pitfalls to avoid in our own organizational structures. Admittedly, however, it is probably unwise to play radical, untested games with the fundamental structure of a large and mature business, only because the losses can be so much larger and more catastrophic. Better to test the idea in a small or medium-sized business first. But it's a good thing for people to experiment generally, and we learned here nonetheless. It would be unfortunate if the only lesson future business leaders took from this episode is never to deviate from the monarchical model of corporate governance.