Andre preaches on the importance of quality, and the dangers of fast and easy compromises.
Although I don't bear the responsibilities and decision-making burden of a CEO (i.e. whether to get a DJ or a magician for the Ping XMas party, blue or green tinted drinks etc), Andre's words did strike a chord for me as an occasional specification writer.
There have been many a time when I felt a specification I was working on could be justifiably called complete, and so get it off my editorial plate. The temptation to finish is strong.
So when ever I feel tempted to quit before the job is done, I think about those spec writers who've come before me.
Did the authors of RFC2119 stop at MUST & SHOULD? I'm sure we could have got by with just those two. But no, 'getting by' wasn't good enough for Bradner et al - they went the extra mile and added MAY as well. And then they added the 'NOTs' as well. That's giving 110%.
As another example, when drafting their charter, did the founders of the WS-Federation TC take the easy path of excluding key input? OK, well bad example.