You don’t need a specialized tool to find your time leaks. You need two weeks of calendar data and these five prompts. Each one is designed to surface a different category of invisible time loss.
Run them in sequence for a complete analysis, or use any single one for a focused look at one category. Total time: under an hour.
Prompt 1: The Structural Pattern Analysis
Use this for: Meeting leaks and calendar fragmentation.
What to give the AI: Your calendar for the past two weeks (export as text or describe meetings by title, duration, and time of day).
Here is my calendar for the past two weeks:
[paste calendar data]
Analyze the structural patterns and tell me:
1. What percentage of my working hours are committed to meetings or scheduled obligations?
2. How many days in the past two weeks had a continuous uninterrupted block of 90 minutes or more available? What time of day did those occur?
3. Do any of my recurring meetings create a fragmentation pattern — placed in a position that destroys the usefulness of the surrounding time?
4. If I wanted to protect one four-hour deep work block each day in the morning, which specific meetings would need to move or be eliminated?
Describe only what you observe in the data. Save recommendations for when I ask.Prompt 2: The Meeting Value Audit
Use this for: Identifying which meetings are justified and which are time leaks.
What to give the AI: Your list of recurring meetings with title, duration, frequency, and your role in each (decision-maker, contributor, or attendance without clear role).
Here is my list of recurring meetings:
[list meetings with title, duration, frequency, and your role]
For each meeting, evaluate:
1. Is there a clear output or decision this meeting produces that couldn't be achieved asynchronously?
2. Is my attendance necessary to achieve that output — or is my role primarily as an information recipient?
3. Based on how you'd evaluate a business case: which meetings have a time cost that appears to exceed their likely value?
Rank the meetings from most to least defensible based on this analysis.Prompt 3: The Context-Switch Cost Estimate
Use this for: Understanding the hidden cost of interruption patterns.
What to give the AI: A description of your notification and communication habits — honest answers to the questions below.
I want to estimate my daily context-switch cost. Please help me calculate it based on the following information about my typical workday:
- How often I check email during work sessions (roughly): [your answer]
- Whether Slack/messaging is open during work: [yes/no and how often I respond]
- Approximate number of meetings per day: [number]
- Whether I have phone notifications on during work: [yes/no]
- How often I switch between applications during a work session: [rough estimate]
Using Gloria Mark's research finding that the average recovery time after a significant interruption is approximately 23 minutes, estimate:
1. Roughly how many significant context switches do my habits above create per day?
2. What is the approximate daily overhead cost in recovery time?
3. Which two or three switching triggers are likely causing the most cumulative damage?Prompt 4: The Micro-Task Load Analysis
Use this for: Identifying whether reactive task processing is fragmenting your day.
What to give the AI: A description of how your communication and task processing currently works.
I want to understand whether my micro-task handling is creating a significant time leak. Here's how my communication and task processing currently works:
- Email: [describe how you process it — continuously, in batches, first thing in the morning, etc.]
- Slack/messaging: [how you respond — reactive as messages arrive, or batched]
- Task requests from colleagues: [how they arrive and how quickly you respond]
- Average number of email/message interactions per day: [estimate]
Based on this, please:
1. Estimate how many individual decision/response events I'm handling per day
2. Calculate the difference in context-switch events between my current distributed approach and a batched approach with two processing windows
3. What is the estimated daily time and cognitive cost of handling these individually rather than in batches?Prompt 5: The Honest Week Review
Use this for: Surfacing leaks that data doesn’t capture — the behavioral and environmental patterns that contribute to time loss.
No data required — just honest reflection.
I'm going to describe last week honestly and I'd like you to identify the patterns that suggest time leaks.
Last week looked roughly like this:
[Write a candid 5-10 sentence description of how your week actually went — not the idealized version. Include: when you felt most productive, when you felt scattered, what pulled you off track, any moments you noticed yourself wasting time, what you checked more than you needed to, what tasks expanded beyond their expected scope.]
After reading this, please:
1. What patterns in my description suggest recurring structural leaks (things that would happen again next week in the same form)?
2. What behaviors am I describing that sound like symptoms of a deeper pattern I'm not naming directly?
3. What single change to how my week is structured would have made the largest difference to how last week felt?After the Prompts: What to Do with the Output
Don’t try to act on everything these prompts surface. That’s how time management projects become their own time leaks.
Take the single finding from each prompt that seems most likely to be costing you the most time. Write those five findings on one page. That’s your rough Leak Map.
Pick the one you can fix this week without needing anyone else’s cooperation. Fix it. Measure the result.
The full framework for turning these findings into a systematic elimination plan is in the Complete Guide to Eliminating Time Leaks with AI.
Start with one prompt. Run it today.
Frequently Asked Questions
-
What data do I need to use these prompts?
Prompts 1 and 2 work with calendar data alone — export your calendar for the past two weeks or describe your typical week in detail. Prompts 3 and 4 work with time-log estimates you can reconstruct from memory in 20 minutes. Prompt 5 works with nothing but honest reflection. Better data produces better analysis, but you don't need perfect records to start.
-
Which prompt should I use first?
Use Prompt 1 first. It provides the broad structural analysis that makes the subsequent prompts more targeted. Running Prompt 3 or 4 without the structural context from Prompt 1 often produces recommendations that address symptoms rather than sources.