How to organize the development process? Best valuation practices?

Question

Business Informatics

What are the best practices for evaluating project/task implementation?
Good day, dear ones.
I have been in web programming for 5 years already and I basically know how to estimate the timing of tasks and projects.
However, when there is a non-trivial task that requires a non-standard approach, I get lost and call the deadline too low as usual, and then suffer from it. I work from home, at night and all that.
I think a similar experience could be not only in my practice. Which of you encountered difficulties in assessing and what methods did you use in order not to deceive yourself or the client?
Also, please advise good literature.
Regards, Ivan.

Answer 1

I have been in the IT field for almost 20 years. And the same trouble. Maybe I just have a little more projects. I can’t move forward until I decompose the tasks into a more or less understandable plan.
More or less understandable is a plan that is detailed to within 4 hours and takes into account my focus factor . Since I am often distracted by various extraneous projects, it is small, somewhere around 0.7-0.75.
At the same time, I have to distribute my attention to several of my active projects, so I estimate how many hours per week I will spend on each of these projects. A week, because in my work this is the best short-term planning period: meetings are often rescheduled, code for related components appears late, etc. So within a week, my labor costs are more stable than on a specific day.
Ultimately, I came to a plan built in MS Project. All tasks for all projects that need to be solved are uploaded to the plan. They are detailed, as I said, into subtasks lasting up to several hours. Connections are being established. The plan is prepared 3-4 months ahead. The baseline is saved. After that, of course, the plan is constantly adjusted. It is fixed every day how much I really spent on each task, new tasks are introduced, some tasks are transferred. In general, it takes half an hour on Monday and 15 minutes on other days.
And I always have in front of my eyes the numbers that I put into the plan initially, the planned indicators as of the moment and the real costs. The analysis of indicators allows me to adjust my estimates when drawing up a future plan.
PS I almost forgot. Managers of various projects, of course, try to load their employees to the fullest. Your plan should not contain overtime! If it is full, then gather managers and let them prioritize your tasks. If everything is important and urgent, then once you can meet good people, but then let them look for your partner.

Answer 2

I always use a simple rule: term = (longest_option * 2) + time_for_testing + time_for_debugging
In non-standard tasks, it is always difficult to estimate the real time of work, so I work only according to this formula.