Wait, but isn't the whole point of training to stop the big mistakes before they happen? If you only give basics, how do you know what someone will think is a small, safe error versus what blows up a client project? That rollout mess sounds like it started as someone's small mistake. Maybe the real trust is in giving people the full picture so they understand how their piece fits.
Last quarter we over-trained a new team and it killed their drive to figure things out. Sometimes people learn more by doing and fixing small mistakes. Your point about cost makes sense, @the_ben, but I've seen the other side where too much hand-holding backfires.
Balance is key. Skipping training sets people up to fail, like in that client rollout mess. But drowning them in info kills any curiosity to learn on their own. The goal should be giving enough basics to prevent big errors, then letting them figure out the rest. It's about trust and letting people make small, safe mistakes. What's the worst training fail you've seen turn out okay?
Read about a tech startup that messed up a launch because they rushed training. The fallout cost them more than a full training program would have. Your experience really drives that home.