A product manager's inbox on a Friday afternoon is a good place to see the difference between urgent and important.
A stakeholder wants a slide updated before Monday. A customer escalation landed in Slack. Someone asks for a quick look at a design. Meanwhile, the research synthesis you've been meaning to finish for two weeks is still sitting there, not urgent at all, and probably the most important thing on the list.
The Eisenhower matrix is a simple way to sort this. Urgent or not, important or not, four boxes. It's old and it's basic. I still find it useful, mostly because it makes me admit what I'm avoiding.
The four boxes, PM edition
| Important | Not important | |
|---|---|---|
| Urgent | Do now: a production incident, a launch blocker, a decision the team is waiting on | Delegate or handle quickly: routine status requests, meeting logistics |
| Not urgent | Schedule: discovery, strategy, writing, reviewing metrics properly | Drop: low-value reports nobody reads, meetings without a clear purpose |
The box that matters most is the bottom left. Important but not urgent. That's where the work that actually changes the product lives, and it's the first thing to get pushed aside when the week gets busy.
How I use it in practice
I don't draw a matrix every morning. But when the week feels out of control, I'll list everything I'm being asked to do and sort it roughly into the four boxes. A few patterns usually show up:
- Something "urgent" turns out to be someone else's deadline with no real consequence if it slips.
- Something important has been sitting untouched because it never had a deadline.
- A few things can be handed to someone better placed to do them.
- A couple of things just don't need doing.
The sorting takes ten minutes. The value is in the honesty.
Ask what makes something urgent
When someone says a request is urgent, I try to ask what happens if it waits a week. Not to be difficult. It's genuinely useful information.
Sometimes the answer is serious: a contract, a customer about to churn, a live bug. Sometimes it's "the VP asked for it." That might still be worth doing quickly, but it's a different kind of urgency. I wrote more about this in the bugs vs features post, where "needed by Monday" often turns out to be a date without a reason.
Protect time for the bottom left
The only way I've found to do important, non-urgent work is to put it in the calendar like a meeting. Two blocks a week for discovery, writing or analysis. If something urgent genuinely needs that time, fine. But it has to justify taking it.
Without that, the important work only happens when there's a lull, and in product there's rarely a lull.
Where the matrix breaks down
It's a rough tool. A few limits:
- It doesn't tell you how big each task is. A small important task and a huge one land in the same box.
- "Important" depends on goals. If the team's goals are unclear, so is everything in the matrix.
- Delegating only works if there's someone to delegate to and they have the context.
For backlog decisions, a framework like RICE is better. The matrix is for your own week.
Try it as a game
Friday 5PM on my portfolio is built around exactly this: tickets arrive and you sort them into do now, schedule, delegate or drop against a timer. The scenarios are fictional and the answers are fixed, so it's simpler than a real week. But it's a quick way to notice your own instinct to treat everything loud as everything important.