Task prioritization is the habit of deciding what to do first by weighing what each task is worth against what it costs, instead of working through whatever is loudest. The method matters less than most guides suggest. What matters is that you pick one criterion, apply it to every task, and keep the result somewhere you will see it. This post compares six common methods on the same task list, shows where each one breaks, and ends with a way to keep the ranking alive on a kanban board.
What is task prioritization, really?
It is two decisions, made in this order. First, which tasks are even candidates. Second, in what order the candidates get done. Most advice jumps to the second and skips the first, which is why people end up with a beautifully ranked list of fifty items that nobody could finish.
So start with a master list: everything you owe somebody, in one place. Then mark the real deadlines, meaning dates somebody else will enforce, not dates you invented. Only then does a method have something honest to work on.
Which task prioritization methods are there?
These are the six that cover nearly every situation. Each gets the same sample list, so you can see how the answer changes with the method:
- Write the quarterly report (due Friday, takes 4 hours)
- Reply to a client complaint (arrived today, 30 minutes)
- Fix the broken invoice template (blocks two people, 2 hours)
- Plan next month's roadmap (no deadline, 3 hours)
- Clear 40 emails (nobody is waiting, 1 hour)
- Learn the new reporting tool (no deadline, 5 hours)
Eisenhower matrix: urgent against important
Sort each task into one of four boxes: urgent and important, important but not urgent, urgent but not important, neither. On the sample list, the complaint and the report go to the first box, the roadmap and the tool to the second, the emails to the third or fourth.
It is fast and needs no numbers. It breaks when everything feels urgent, so the first box swallows the list. The full setup is in the Eisenhower matrix on a kanban board.
MoSCoW: must, should, could, won't
Label each task Must, Should, Could or Won't, for a fixed window such as this week. The report and the complaint are Musts, the invoice template a Should, the roadmap a Could, the emails and the tool Won't, for now. MoSCoW comes from software project planning, and Wikipedia has a short history of it.
Its strength is that it forces a "no". Its weakness is that it does not rank inside a label, and Musts tend to multiply.
Impact and effort: the quick-wins grid
Score each task low or high on impact and on effort. High impact with low effort goes first. The invoice template wins here: two hours, and it unblocks two people. The tool and the roadmap are high effort, so they wait.
This is the best method for choosing between improvements. It is the wrong one for deadlines, because a high-effort task due Friday still has to be done on Friday.
RICE: a score instead of a feeling
RICE multiplies reach and impact and confidence, then divides by effort. It gives you a number per task, which is useful when you must defend a ranking to other people. For a personal list it is overkill: you will spend longer scoring than the emails would take to clear. Use it for product backlogs with many stakeholders, and skip it for your own week.
ABCDE: letters with numbers inside
Give every task a letter, A for must do through E for eliminate, then number the tasks within each letter. The point is the second step. Where MoSCoW stops at a label, ABCDE makes you choose between two A tasks. On the sample list that means deciding whether the report (A1) comes before the complaint (A2) or after, and writing down why.
Ivy Lee: six tasks, ranked the night before
Each evening, pick the six tasks for tomorrow, rank them, and work through them in order. It needs no criterion at all beyond your own judgment, which is its strength and its flaw. The Ivy Lee method on a kanban board covers the routine and where it fails.
How do you choose a method?
Choose by the shape of your problem, not by which method sounds the most rigorous.
| Your situation | Use this |
|---|---|
| Too many things feel urgent | Eisenhower matrix |
| You say yes to everything | MoSCoW, with a hard cap on Musts |
| Choosing between improvements | Impact and effort |
| Defending a ranking to others | RICE |
| Two important tasks tie | ABCDE |
| You lose the start of every day | Ivy Lee |
You can also stack two. A common pairing is Eisenhower once a week to sort the pool, then Ivy Lee each evening to pick from it. Do not stack more than two, because the sorting starts to take longer than the work.
Where do prioritization methods fail?
Three failures account for most abandoned systems.
The ranking goes stale. You sort on Monday, an urgent request lands on Tuesday, and by Wednesday the list describes a week that no longer exists. A method with no re-sort step is a one-off exercise, not a system. The list is too long to rank. If you have 60 open tasks, ranking all of them is the wrong first move. Cut first, rank second. Anything you would not miss if it vanished goes to a "someday" list or gets deleted. Everything gets the top label. If more than a third of the list lands in the first box, the criteria are too soft. Tighten them: a task is only a Must if something concrete breaks without it, and only urgent if a person is waiting.How do you keep priorities visible on a kanban board?
A board turns a ranking from a document into a place. The setup is small.
Make four columns: Backlog, Next, Doing and Done. Everything lives in Backlog, unranked. Next holds the tasks you have committed to for this week or day, in priority order, top card first. Doing holds what you are working on right now.
Then apply two limits, as covered in WIP limits explained. Cap Next at around six cards, so you cannot commit to more than you can finish. Cap Doing at one or two. The cap on Next is what makes the prioritization method matter, since a full column forces you to drop something before adding something.
When an urgent task arrives, put it in Next where it belongs by your criterion, and if the column is now over its limit, push the lowest card back to Backlog. The cost of the interruption is visible, not hidden.
Re-sort once a week in a weekly review: empty the old Next, pick the new one from Backlog using your chosen method. That single habit solves the stale-ranking problem. A free board on EasyKanban's minimal kanban page is enough to try the four-column layout without an account.
FAQ
What is the best task prioritization method?
There is no single best one. The Eisenhower matrix suits people who feel everything is urgent, impact and effort suits choosing between improvements, and Ivy Lee suits people who lose the start of the day. Pick the one that matches your problem and use it for a month.
How do you prioritize tasks when everything is urgent?
Ask who is waiting and what breaks if the task slips. If nobody is waiting and nothing breaks, it is not urgent. Then rank the truly urgent tasks by effort, shortest first, so you clear the queue fast. If more than a third still look urgent, the deadline itself needs a conversation.
How many priorities should you have at once?
Fewer than feels comfortable. For a day, three to six tasks is realistic. For a week, around six committed tasks plus whatever is in progress. More than that and the list stops being a priority list and turns back into a backlog.
Should you prioritize by deadline or by importance?
Both, in sequence. Deadlines set what must happen and when, so handle those first. Importance then orders the tasks that have no deadline, which is where the Eisenhower matrix helps: the important but not urgent box is where long-term work lives and where it usually gets squeezed out.
How often should you re-prioritize?
Once a week for the full list, and briefly each day for the top few tasks. A weekly review catches drift before it grows. Re-sorting every hour is a sign the criteria are unclear, not that the work has changed.