Making yourself easy to be scheduled: Part 5 - Padding leads to the dark path
During the earlier section on the daily logging exercise, you probably thought to yourself, “Why should I go through that hassle when what I really need to do is pad my estimates so I will have extra time to get tasks done?” That is a very fair question. Think about what you are doing when you pad your estimates, though.
In absence of teaching yourself how to make good predictions through the daily exercise described in the earlier section (or some other method), any other estimate is going to be a guess based on your memory of your experiences. The padding you add to that is essentially another guess, designed to cover the inconsistencies of the first guess. The synthesis of two uneducated guesses does not give you a more accurate estimate. Instead, it leaves you with the expectation that you have no idea what you are talking about to begin with, and your promise to your scheduler/boss is based on little more than whim. Do you really want your next raise determined by the correctness of multiple guesses added together?
Admittedly, an estimate is a fancy word for a guess that implies greater accuracy. The daily logging exercise is meant to make your uneducated guesses educated, thereby turning them into estimates. Padding is essentially an excuse to make up some arbitrary number of scheduling wiggle room to account for the inaccuracies in your estimation process. If you know enough about yourself and your own experience with similar tasks to begin with, your original estimate will be good enough to not need the extra padding in case you are wrong.
Instead of padding, the better approach is to isolate the sub-parts of your tasks for which there are unknowns, such as those dealing with technologies that you have not used before, into their own tasks. Estimate them as best you can, but be honest with your scheduler that there are things about these particular items you simply cannot predict with an acceptable amount of accuracy. These tasks are ill-defined and needs to be watched carefully. This signifies to the people in power that these pieces of work may be risky and lets them make more informed decisions accordingly. When you make this discovery while the schedule is under way, it is important to follow this same set of steps and clearly communicate the situation (and any alternative approaches) to your scheduler as soon as possible.
Usually when a task is unpredictable, it is because not everything surrounding it is a known quantity. When possible, try to learn more about the unpredictable task before starting it so that there are fewer unknowns to deal with before you spend too much more time on it. Sometimes that may not be feasible for scheduling reasons, but when it is it can help you increase the accuracy as much as possible without relying upon padding to somehow save the day.
In absence of teaching yourself how to make good predictions through the daily exercise described in the earlier section (or some other method), any other estimate is going to be a guess based on your memory of your experiences. The padding you add to that is essentially another guess, designed to cover the inconsistencies of the first guess. The synthesis of two uneducated guesses does not give you a more accurate estimate. Instead, it leaves you with the expectation that you have no idea what you are talking about to begin with, and your promise to your scheduler/boss is based on little more than whim. Do you really want your next raise determined by the correctness of multiple guesses added together?
Admittedly, an estimate is a fancy word for a guess that implies greater accuracy. The daily logging exercise is meant to make your uneducated guesses educated, thereby turning them into estimates. Padding is essentially an excuse to make up some arbitrary number of scheduling wiggle room to account for the inaccuracies in your estimation process. If you know enough about yourself and your own experience with similar tasks to begin with, your original estimate will be good enough to not need the extra padding in case you are wrong.
Instead of padding, the better approach is to isolate the sub-parts of your tasks for which there are unknowns, such as those dealing with technologies that you have not used before, into their own tasks. Estimate them as best you can, but be honest with your scheduler that there are things about these particular items you simply cannot predict with an acceptable amount of accuracy. These tasks are ill-defined and needs to be watched carefully. This signifies to the people in power that these pieces of work may be risky and lets them make more informed decisions accordingly. When you make this discovery while the schedule is under way, it is important to follow this same set of steps and clearly communicate the situation (and any alternative approaches) to your scheduler as soon as possible.
Usually when a task is unpredictable, it is because not everything surrounding it is a known quantity. When possible, try to learn more about the unpredictable task before starting it so that there are fewer unknowns to deal with before you spend too much more time on it. Sometimes that may not be feasible for scheduling reasons, but when it is it can help you increase the accuracy as much as possible without relying upon padding to somehow save the day.
- Part 1, Introduction
- Part 2, The Lego Exercise
- Part 3, Teaching yourself better estimation
- Part 4, Understanding your scheduler
- Part 5, Padding leads to the dark path
- Part 6, Be good to yourself and your teammates
- Part 7, Final Thoughts
Labels: Reference
0 Comments:
Post a Comment
<< Home