One challenge I've found is how to fail and learn without ruining plans and losing trust. In order to write up a plan you have to make assumptions. As with any assumption, you can easily get it wrong, throwing your well organized plan into a tailspin. When your assumptions impact commitments to others, it is natural that those folks might lose trust that you'll deliver what you say. At the same time, what about failing fast and learning from those failures?
There is not a clear answer to this that I've found. If you know, please tell me! There are a couple things that seem to be reasonable tactics.
One thing is to always be setting expectations. If you are starting with a wild guess, you need to communicate to those folks paying attention that this is a shot in the dark. If your guess was way off, you need to follow up quickly with those folks to let them know things changed because your guess was wrong. I imagine that as you develop a reputation for hitting your targets, that communication becomes less important for those commitments you are likely to hit. But for the tasks that require guesses, you're still on the hook to over communicate.
Another thing to do is call out your failures and reshape the narrative. Failure is a great way to learn, but if you don't clarify what you learned, someone watching won't know that there was some value created within that failure. For example, lets say you try some new UI on a website and it reduces traffic by 30%. That data is valuable. It is saying don't do whatever the UI change was doing. If we fail and we don't communicate what we learn and how that learning is valuable, then we leave it up to others to make assumptions about our performance. It is better to take ownership of a negative outcome and codify what didn't work so you don't make the same mistake.
The big takeaway is to communicate. Our instinct is to try and hide the failure or see how you can some how manipulate a failure into a win. Instead, it is best to be honest, up front and communicate what was learned. If we also want to create a culture of learning, we need to communicate that things may fail. Make it clear you're guessing and things could fail completely.
I've seen this practice work when writing music. We have written songs and someone wants to try some idea they hear. When we were first starting out, someone would say, "Let's do this part here" and we'd all try it out. The problem was that the person didn't leave themselves an out. If the part didn't work, the person who had the idea needed to drop it and it was harder because of the way they presented it. Now when we write songs, new ideas are prefaced with phrases like "This may not work, but what if we did this?". We can try it and if it isn't working, we drop it. What's more, we often say why it isn't working and learn something. For example, it might make another part lose some impact or the notes are higher and it doesn't sound as full. We start to get a shared understanding of what we like about a song and we improve because we are naming things we hear. More generally, it makes it safe to experiment, fail and learn.