Planning

Estimation

Estimation is a very hard process, especially for a creative process where the thing we’re building has typically never been built before. However, through experience we can get slightly better at knowing whether some type of work may be straightforward or if requirements are overseen and thus more complicated than was expected.

So to still be able to estimate somewhat how complex certain tasks are relative to others, we use estimation values that inherently don’t relate to time, just complexity; a.k.a. “story points”. ZenHub includes a feature to keep track of these.

This information can then be used to see what our current velocity is, so that we might have a better insight into whether or not we’re on-schedule.

To come up with these estimates, we use a tool called Planning Poker. Essentially:

1. we all come up with estimates for each story in the Estimation pipeline on our own (this should just take a minute max per story)
2. during a planning meeting we go through our estimates for each of the stories
3. we hear arguments for the outliers (i.e. the lowest and the highest estimate)
4. we pick the estimate that we all agree on

Prioritisation

In order to ensure that we do actually ship new features at some point, e.g. the end of a sprint, it is important that we prioritise certain requirements of a feature over others.

Maybe requirement X has a very high estimated story point, but is not as essential to the feature as a whole as other requirements are. Thus it should have a lower priority compared to those that are essential, so that the feature may still be shipped, even if requirement X did not get completed.

Prioritisation is done by the designer and engineer that own a feature and after estimation.