Where the day really goes

Being busy and getting results are different things

Almost everyone who complains about lacking time already works hard. Speed is not the problem. The problem is that the day breaks into fragments, and nothing that requires thinking fits inside a fragment.

Before building any system, you need a fact rather than a feeling. The fact is this: how many uninterrupted stretches longer than forty minutes exist in your day. For most people with a work inbox and a couple of messengers, the answer is zero or one.

Three sources of loss

Switching. Returning to the previous depth after an interruption takes 10 to 23 minutes, a figure from Gloria Mark's research. Five interruptions eat the whole morning even though each one lasted thirty seconds.
Standing by for incoming messages. Not the conversation itself, but the background readiness for it. You are neither resting nor working.
Undefined tasks. The item "sort out the report" has been sitting on the list for two weeks because the first step is unclear. Every glance at it costs attention and moves nothing.

Measure instead of guessing

For three working days, write down what you are doing every time you change activity, on the actual switch rather than on a timer. Keep the format minimal:

09:10 email
09:25 report
09:31 chat, question from a colleague
09:38 report
09:44 chat
10:05 call

After three days, hand the log to an AI with this request: "Here is a three-day log. Count the number of switches per day, the average length of an uninterrupted stretch, the longest stretch, and the three activities that consumed the most time. Do not give advice." The ban on advice matters — otherwise you get a list of platitudes instead of your own numbers.

People running this measurement for the first time usually find 30 to 60 switches per day and an average stretch of 8 to 12 minutes.

Insight. The issue is not the volume of work but how it is sliced. Eight hours in ten-minute pieces produce less than three uninterrupted hours, long tasks require holding context in your head, and it is dumped at every interruption.

Common mistake. Installing a system before measuring. Someone buys a planner, configures tags and categories, and abandons it a week later — because the real problem was six back-to-back meetings, and no planner solves that.

Pro tip. Count not how long you worked but how many times you started over. The second number explains your fatigue better than the first, and it responds to change faster.

Cheat sheet

Facts first, system second.
Three days of activity-switch logging is the minimum honest measurement.
Key numbers: switches per day and average stretch length.
The AI counts from your log but gives no advice at this stage.

1. How long does research say it takes to return to previous depth after an interruption?

2. Why forbid the AI from giving advice during the measurement stage?

3. What signals that a task is stuck because it is undefined?

🔒 Answer the question correctly to move on to the next lesson.