> however, I was scared sh1tless giving any kind of projection on any kind of project with me as the sole developer doing things I've never done before; it seemed to me to be absurd.
I'm long past my junior days and this still bugs me. In case it helps anyone else, now I give my estimations with two [low, mid, high] qualifiers next to them:
1. Uncertainty: this is where I signal whether I think the task is well defined and we have the expertise to deal with it or not. The estimation gets a multiplier (however stupid I think the task is) when this is not low, and I make it clear that uncertainty can be reduced by either better defining the task or breaking it into an "exploratory" phase (where we try to come up with a tech-demo to get a better "feel" of that new thing) and then an execution phase that will turn it into a reality (which will be estimated _after_ the exploratory one is done).
2. Environmental risk: this signals how external factors can impact the task (e.g.: the client is complicated to work with, this necessitates collaboration from a third company that may or may not be fluid, we require key people from the team that is busy doing something else, etc.)
I've found that this both reduces my anxiety and tunes the client's expectations much better (it becomes much easier to explain after-the-fact why two tasks that had a similar estimation ended up taking wildly different efforts).
I'm long past my junior days and this still bugs me. In case it helps anyone else, now I give my estimations with two [low, mid, high] qualifiers next to them:
1. Uncertainty: this is where I signal whether I think the task is well defined and we have the expertise to deal with it or not. The estimation gets a multiplier (however stupid I think the task is) when this is not low, and I make it clear that uncertainty can be reduced by either better defining the task or breaking it into an "exploratory" phase (where we try to come up with a tech-demo to get a better "feel" of that new thing) and then an execution phase that will turn it into a reality (which will be estimated _after_ the exploratory one is done).
2. Environmental risk: this signals how external factors can impact the task (e.g.: the client is complicated to work with, this necessitates collaboration from a third company that may or may not be fluid, we require key people from the team that is busy doing something else, etc.)
I've found that this both reduces my anxiety and tunes the client's expectations much better (it becomes much easier to explain after-the-fact why two tasks that had a similar estimation ended up taking wildly different efforts).