As your https://en.wikipedia.org/wiki/Cone_of_Uncertainty gets increasingly confident, the key is to always think at least two steps ahead and take seriously what your estimates are, esp. long poles in the process.
I was an engineer for 20 years. Then an engineering manager for two years. And then I got hit with my first big project that was a month delayed due to dependencies we didn't anticipate.
The big learning there finally was always work backwards from your dates and ensure high confidence. If you're not high confidence in your date, you need to embed yourself with your dependencies and own the dependency outcomes until you can vouch for all of them (esp. your long poles, i.e., the dependencies that will take the longest or have the highest risk at particular stages in the project), ensure they are all high confidence.
There's no other way. Wisdom is knowing what to emphasize the next time in the planning or design phases.
When you're running late, the best thing you can do is begin that process and figure out what it really takes to get to a high confidence date (assuming no extra hours worked that risk burning out you, the team, etc.). Your stakeholders, if they accept a moved date, will want you to explain why you are high confidence for the new date given the trust you've lost for the previous date. You'll have to earn that trust back via repeated on-time delivery (or via raising the quality bar or providing other value to your stakeholders).
I was an engineer for 20 years. Then an engineering manager for two years. And then I got hit with my first big project that was a month delayed due to dependencies we didn't anticipate.
The big learning there finally was always work backwards from your dates and ensure high confidence. If you're not high confidence in your date, you need to embed yourself with your dependencies and own the dependency outcomes until you can vouch for all of them (esp. your long poles, i.e., the dependencies that will take the longest or have the highest risk at particular stages in the project), ensure they are all high confidence.
There's no other way. Wisdom is knowing what to emphasize the next time in the planning or design phases.
When you're running late, the best thing you can do is begin that process and figure out what it really takes to get to a high confidence date (assuming no extra hours worked that risk burning out you, the team, etc.). Your stakeholders, if they accept a moved date, will want you to explain why you are high confidence for the new date given the trust you've lost for the previous date. You'll have to earn that trust back via repeated on-time delivery (or via raising the quality bar or providing other value to your stakeholders).