Parallel tasks, focus, and shorter cycles

Ozan Onay — 02 Apr 2011

Caleb Elston writes of the culture of “serial”-tasking at Yobongo, raising the point that when we attempt to handle many tasks in parallel, we may irrational elect to work towards low priority or low yield tasks:

While there are indeed many things to get done, there are usually only one or two that are critical at any point in time. The opportunity cost of sacrificing attention on the high priority tasks just to feel like you are 'getting more done' is economically irrational.

When deciding on product release scheduling or iteration length, two commonly citied reasons for shorter cycles are increased accountability of the team and more timely feedback from external stakeholders. I would add that shorter iterations force a team to restrain the sheer number of tasks they hope to complete within the time period, reducing the temptation to knock off small, simple tasks rather tackling the most important.

While we might hope that assigning priorities to tasks is sufficient to effectively prioritize tasks, this suffers from the restraint bias: we are likely to underestimate the restraint required to continue working on an important but difficult task when we’ve also decided that smaller, easier tasks are within scope. It is better to avoid the necessity for restraint by only planning to complete a small number of difficult tasks (preferably just one!) knowing that our tendency to seek easy wins will cover off the easy tasks at some stage. I’ve found that shorter cycles, by reducing optimism bias and the effect of the planning fallacy, help limit scope and thereby reduce the costs of multitasking.