The paradox of estimating
'Individuals and interactions over processes and tools' - this is a key component of the agile manifesto. As teams, especially in bigger orgs, we are susceptible to process creeping over everything.
Process often emerges as a reaction to a failure or to a sense of a lack of control. While process isn't bad in itself, it shouldn't override or completely replace natural interactions between people.
One particular challenge teams face when dealing with process is when any talk of estimation comes about. As delivery people, we often want to introduce a way of understanding time scales and effort so that the team can use this information to better itself and improve working practices.
However, estimation in itself can become quite an onerous process which can undermine the commitment to human interactions > process. This is especially since it isn't always clear what the value of complex estimations are.
That's why I think it is best to keep estimating simple in teams and organisations.
In teams, it makes sense to speak the language of the team so that people can estimate or order tasks (if this is necessary at all) in a human way.
Different teams will need and demand different ways of doing things however I quite like the light approaches that the following estimating methods offer:
- Lunar Logic's No Bullshit estimation cards
- Swimlane sizing
- Counting stories (focus on breaking them down to similar sizes)
Team's will often reach a point naturally where they are sizing items similarly (into manageable items) and so it doesn't usually make sense to enforce a complex system of sizing when they have already done the work to minimise discrepancies between sizes (which are often not representative after doing them anyway).
Keep it simple and estimating doesn't need to feel like a process.