Starting · cornerstone
What a deep-work ritual actually looks like
The timer is only the middle. A repeatable focus session starts before you press play and ends by making the next session easier to begin.
Most pictures of deep work start with the timer.
I think the timer is the middle.
The hour starts before you press play
When I used to think about a focus session, the recipe was simple: sit down, put on music, start 45 minutes, work.
That part is useful. It is just incomplete.
Because some of my worst “focus” sessions started with a timer and no real decision about what I was trying to finish. And some of my most productive ones ended with me so mentally done that starting again the next day felt ridiculous.
That is why Ember is built around four parts instead of one: Connect, Focus, Recover, Return.
Enter the work. Protect the work. Leave the work. Make the next entry easier.
1. Connect: decide what this block is for
Before you open another tab, answer one question:
What will be different when this block ends?
Not “work on the website.” That can mean anything.
“Rewrite the hero section.” “Finish the first draft.” “Reconcile these three accounts.” “Study chapter four and make ten flashcards.”
Then name one thing you are not doing until the block is over. Inbox. Phone. Researching a completely different problem that suddenly became fascinating the second the real work got uncomfortable.
There is good research behind planning in advance. Implementation-intention studies show that specific if-then plans can make goal follow-through more likely. Ember’s prompt is not identical to every protocol in that research, so I do not call the feature “proven.” The useful idea is simpler: make fewer decisions after you start.
2. Focus: one meaningful thing
This part is boring on purpose.
One task. One protected block. As few decisions as possible about where your attention should go next.
Research on attention residue helps explain why. When people switch away from unfinished work, part of their attention can stay with the previous task, and performance on the next task can suffer.
That is why a “quick check” is not always quick.
The two-minute message might take two minutes to answer. The mental switch can hang around longer.
Timers, music, blockers, full-screen modes: this is where those tools earn their place. They help hold the room steady once you already know what the room is for.
3. Recover: stop giving your brain another job
When the timer rings, the instinct is usually to reward yourself with more input.
Open the phone. Check messages. Watch something. Scroll “for five minutes.”
Technically, you stopped working.
But did you recover?
Research on micro-breaks suggests short breaks can help people feel less fatigued and more energized, even though the evidence that they improve overall performance is mixed. Other work on micro-break activities suggests the kind of break matters too.
So I prefer a very simple rule: for a few minutes, reduce input.
- Stand up.
- Look somewhere farther away than your screen.
- Get water.
- Breathe more slowly.
- Walk to another room.
- Draw something badly enough that it cannot become another project.
You are not trying to become enlightened in five minutes. You are just changing state.
4. Return: do your future self one small favour
Before you close the session, write down three things:
- What did I finish?
- What is still open?
- What is the first step when I come back?
This takes almost no time, which is probably why it is so easy to skip.
But an unfinished task is much easier to return to when you do not have to reconstruct the entire mental state from yesterday.
The evidence around unfinished tasks is more complicated than the internet version of the Zeigarnik effect usually suggests, so I would not turn this into a grand scientific claim. I just know the practical difference between opening a project and thinking “where was I?” versus seeing one sentence that tells me exactly what comes next.
A 60-minute Ember-style ritual
Here is one way to use an hour. Change the times if they do not fit you.
| Phase | Time | What you do |
|---|---|---|
| Connect | 2 min | Write the finish line for this block and one thing you are not doing. |
| Focus | 45 min | One task. If another thought appears, park it on paper and return. |
| Recover | 8 min | Leave the screen. Reduce input. Breathe, move, look away, or walk. |
| Return | 5 min | Write: finished, still open, first step next time. |
The exact minutes are optional.
The sequence is the useful part.
What this ritual cannot do
It cannot make hard work easy.
It cannot replace sleep.
It cannot turn every session into flow.
And none of the studies above proves that Ember “rewires” your brain, boosts your IQ, or unlocks a secret attention state.
Good. We do not need any of that.
We need a repeatable way to sit down, know what matters, protect the block, recover, and make it easier to return.
That is enough.
Sources, and their limits
- Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493–503.Shows that 'if–then' plans about when and where to act raise the rate of actually acting. It does not show anything about the quality of the work once started.
- Leroy, S. (2009). Why is it so hard to do my work? The challenge of attention residue when switching between work tasks. Organizational Behavior and Human Decision Processes, 109(2), 168–181.Lab studies with students: leaving a task unfinished, or under time pressure, left more of it in mind during the next task and lowered performance on it. Effects were measured over minutes, not days.
- Zeigarnik, B. (1927). Über das Behalten von erledigten und unerledigten Handlungen. Psychologische Forschung, 9, 1–85.The original finding that interrupted tasks are remembered better than completed ones. Replications since have been mixed; treat it as a useful description, not a law.
- Bennett, A. A., Gabriel, A. S., & Calderwood, C. (2020). Examining the interplay of micro-break durations and activities for employee recovery. Journal of Occupational Health Psychology, 25(2), 126–142.Field data on short breaks during the workday: breaks that reduce input help recovery more than breaks that add it. Self-reported, one workplace sample.