How I plan my week as an engineering manager

Engineering manager weeks are unusually easy to overbook. A project needs a decision, a report needs feedback, one-on-ones fill the calendar, and the deep work keeps getting pushed to an imaginary quiet afternoon.

I do not plan the week to control every hour. I plan it to see the tradeoffs early. The output is a forecast: a small set of outcomes, the time they require, and enough open space for the work I cannot predict.

1. Review last week before planning the next one

I start with three questions: What worked? What did not? What will I change?

This keeps the weekly plan connected to reality. If a two-hour technical review took four hours last week, I should not copy the same estimate into a new calendar. If afternoons disappeared into small requests, the answer may be a protected morning block, not more ambitious goals.

I also carry forward unfinished work deliberately. An incomplete task is not automatically important just because it survived the week. I either schedule it, delegate it, reduce it, or drop it.

2. Put personal foundations in first

Before project work, I mark the non-work foundations that keep the rest of the plan honest: sleep, meals, exercise, family time, and any personal commitment that cannot move.

This is not a wellness garnish. Those hours already exist whether the calendar admits it or not. Planning work first and life second creates a fictional capacity number, then turns normal human needs into interruptions.

If the plan only works when sleep, meals, and family go perfectly, the plan does not work.

3. List the commitments already in motion

Next I collect what the week is already asking for. I scan the calendar, active projects, open tasks, team promises, and anything waiting on my decision.

I group them loosely rather than building a complicated taxonomy:

• People: one-on-ones, feedback, hiring, coaching, and difficult conversations.
• Delivery: decisions, reviews, unblockers, and project risks.
• Deep work: writing, analysis, design, or technical thinking that needs focus.
• Operations: planning, reporting, admin, and recurring meetings.

The list is an inventory, not the plan. Its job is to expose demand before I choose what deserves capacity.

4. Choose a few goals and one must-win outcome

I choose a small number of weekly goals, then name one must-win outcome. The must-win is the result that would make the week meaningfully successful even if reactive work consumes more time than expected.

It needs to be an outcome, not a vague category. “Strategy” cannot be finished. “Send the decision memo with a recommendation” can. “Support the team” is endless. “Give feedback on the rollout plan before Thursday” is concrete.

5. Put the work into actual hours

A task list hides capacity. The calendar exposes it. I place the must-win work into the week first, ideally during the hours when I can think most clearly. Then I place meetings, one-on-ones, reviews, and the smaller tasks that have genuine deadlines.

I do not fill every open block. Engineering management contains interrupts by design. A production problem, a people issue, or a decision that arrives late can matter more than the original plan. Open time is how the week absorbs that reality without collapsing.

For focus work, I write the intended output in the block. “Deep work” is easy to ignore. “Draft the first two sections of the hiring plan” tells me how to begin.

6. Scale back before Monday makes the decision for you

Once everything is visible, I look for the lie. Usually the draft assumes that every meeting ends on time, every task takes the optimistic estimate, and no one needs help.

I remove, defer, delegate, or shrink work until the week has breathing room. This is the most valuable step because it turns prioritization from a slogan into a visible tradeoff.

During the week, I treat the plan as a forecast, not a prison. If priorities change, I revise it. The goal is not to prove that Sunday's guesses were perfect. The goal is to keep the current tradeoffs clear.

Copy the weekly planning template

The template is plain Markdown so it works in Ariv, Obsidian, Logseq, VS Code, or any text editor. Keep it short enough to finish every week.

# Weekly plan: YYYY-MM-DD

## Review last week
- What worked?
- What did not?
- What will I change?

## Foundations
- Sleep:
- Meals:
- Exercise:
- Family and personal commitments:

## Existing commitments
- People:
- Delivery:
- Deep work:
- Operations:

## This week's outcomes
- Must-win:
- Goal 2:
- Goal 3:

## Time blocks
- Monday:
- Tuesday:
- Wednesday:
- Thursday:
- Friday:

## What I am not doing
- Deferred:
- Delegated:
- Dropped:

## Friday review
- Result:
- Lesson:
- Carry forward:

A short weekly planning checklist

1. Review what actually happened last week.
2. Protect sleep, meals, exercise, family, and fixed personal time.
3. Collect existing commitments from the calendar, projects, and task list.
4. Choose a few outcomes and one must-win.
5. Block the work into real hours.
6. Leave open capacity for management work you cannot predict.
7. Scale back until the plan is believable.

Research sources

• Gollwitzer and Sheeran: implementation intentions and goal achievement, a meta-analysis of 94 tests.
• Masicampo and Baumeister: Consider It Done!, five studies on plans and the cognitive effects of unfinished goals.