Four ways to prioritise tasks and optimise productivity

At 7:45am on 12th March 2023, a commuter train derailed near Birmingham, having collided with an abandoned vehicle on the track. The rush hour crash left 53 passengers injured, ranging from minor wounds to life threatening injuries. Emergency services rushed victims to hospital where doctors had to prioritise treatments with extremely limited resources.

Using a triage system, medical teams categorised patients based on severity and urgency:

1. High Priority (Red Tag): Patients with severe but treatable injuries, like internal bleeding and collapsed lungs, were treated immediately to maximise survival.
2. Medium Priority (Yellow Tag): Those with serious but non-life-threatening conditions, like fractures and burns, were stabilised and treated later.
3. Low Priority (Green Tag): Patients with minor injuries received first aid and waited until critical cases had been handled.
4. Unsurvivable (Black Tag): Patients with catastrophic, untreatable injuries were deprioritised to focus resources on saving others.

By applying this weighted processing strategy, medics maximised survival rates: 49 of the 53 injured passengers recovered.

Choosing the right productivity metric

The metric you choose shapes the behaviour you get. - Clayton Christensen

If we plan to complete all tasks on a list then any ordering of them will take the same amount of time. Hence, to suggest one approach to task scheduling is better than any other, we must decide what we are trying optimise. Productivity metrics to choose from include:

1. Deadline compliance (Earliest due Date),
2. Avoid overload (Moore’s Algorithm),
3. Get things done (Shortest Processing Time),
4. Prioritise importance (Weighted Processing Time).

1. Deadline compliance (Earliest Due Date)

Deadlines force you to make tough decisions, but they also make you focus. – Seth Godin

Deadlines are often the key factor in scheduling tasks with lateness determining their urgency. The best strategy to minimise maximum lateness (across all tasks) is the Earliest Due Date approach. Complete the task with the nearest deadline first. Task lengths are irrelevant; only due dates matter. Prioritising time-sensitive tasks reduces the risk of missed deadlines.

2. Avoid overload (Moore’s Algorithm)

You can do anything, but not everything. - David Allen

When minimising the number of overdue tasks is more important than reducing lateness, Moore’s Algorithm provides a suitable modification to the Earliest Due Date strategy. When deadlines cannot all be met, discard the most time consuming task to maximise on-time completions. This approach applies beyond scheduling, encouraging prioritisation by eliminating unmanageable workloads, aligning with the productivity principle of saying no to less critical tasks.

3. Get things done (Shortest Processing Time)

The beginning is half of every action. - David Allen

To complete tasks quickly, the Shortest Processing Time strategy is ideal. It prioritises the shortest task first, minimising total completion time and rapidly reducing the number of outstanding tasks. This alleviates cognitive load by making workloads feel more manageable, sustaining momentum.

4. Prioritise importance (Weighted Processing Time)

Nothing is less productive than to make more efficient what should not be done at all. - Peter Drucker

Tasks are not of equal importance. The Weighted Processing Time strategy prioritises tasks based on their value divided by duration, completing those with the highest value-per-time ratio first. This is the approach I apply by default. In my corporate job, I prioritise revenue (or profit) per unit time maximising tasks. In relation to personal finance, I paid off credit cards with the highest interest rates first (debt avalanche method).

Other resources

Getting Things Done talk by David Allen

Debugging Productivity post by Phil Martin

Make Time post by Phil Martin

Brian Christian suggests, Effective scheduling is about implementing the best process, not just focusing on results.

Have fun.

Phil…