5 AI Prompts to Analyze Your Focus Data (Copy and Use Today)

Most people who want to improve their focus have a vague sense that something is wrong — too many interruptions, too little deep work, sessions that end before they should — but no systematic way to see it. Subjective feelings of productivity correlate poorly with actual output. The problem is not that you lack intuition about your focus; it is that intuition compresses a week into a single impression and drops the specifics that would tell you what to actually change. A session log combined with a structured AI prompt does something different: it holds the raw data, surfaces patterns across sessions, and can tell you whether Tuesday at 9 a.m. consistently beats Thursday at 2 p.m. — and by how much. These five prompts work through that analysis systematically, from a first-time baseline read to a month-level trend review. The one constraint: they require real data. Feelings alone will not give the AI enough to work with.

These prompts assume you have a basic session log: date, task type, session length, distraction count, and quality rating. Paste your log after each prompt. If you do not have data yet, the first prompt will tell you what to start collecting.

Prompt 1: First-Time Baseline Analysis

Use when: You have one week of session data and want a structured first read of what the numbers actually show.

Here is my first week of focus session logs. Each entry includes:
date, start time, end time, task type, distraction count during the session,
and a quality rating (1=poor, 2=adequate, 3=excellent).

[paste log]

Please calculate:
1. My average deep hours per day
2. My session completion rate (sessions that ran to planned end time)
3. My average distraction count per hour

Then tell me: which of these three numbers shows the most room for improvement,
and what one question I should be asking about my focus environment based on
this data.

What good output looks like: The AI names one specific metric as the priority improvement area — not all three — and the question it raises points at a testable environmental variable (time of day, task type, meeting load) rather than a general suggestion to “reduce distractions.” If the response lists all three metrics as equally important or recommends habits without reference to your data, the prompt did not have enough specificity. Re-paste with clearer labels on your quality ratings.

Prompt 2: What Conditions Drive My Best Sessions?

Use when: You have two or more weeks of logged sessions and want to know what environmental or scheduling factors correlate with your best work — not what you think drives them, but what the data shows.

Here are [X] focus session logs from the past [X] weeks.
I've also noted weekly context: sleep quality (1–3 scale) and number of
scheduled meetings per day.

[paste log]

Please identify:
1. The two conditions most consistently associated with my quality-3 sessions
2. The two conditions most consistently associated with my quality-1 sessions
3. Whether my distraction count appears to differ by time of day, task type,
   or any other variable you can see in the data

Keep your analysis grounded in what the data actually shows. Flag anything where
the pattern is weak or based on only a few data points.

What good output looks like: Two specific conditions — not categories like “good sleep” but something like “sessions starting before 9 a.m. with zero scheduled meetings in the first two hours” — tied to your quality-3 data points, with a clear flag on where the sample is too small to trust. A failed response over-interprets two or three data points as a clear pattern without noting the sample size limitation.

Prompt 3: Diagnosing a Bad Week

Use when: A specific week’s metrics dropped noticeably below your recent baseline and you want hypotheses about the cause before you attempt a fix.

My focus metrics dropped significantly this week compared to my recent baseline.
Here is this week's session log and my previous two weeks for comparison:

[paste all three weeks]

My best guess at what was different this week: [briefly describe — more meetings,
new project started, stressful deadline, disrupted sleep, etc.]

Please:
1. Confirm whether the metrics drop is as significant as it felt
2. Identify what specific metrics dropped most (volume, completion, distractions)
3. Given what I described as different, generate two or three hypotheses about
   what caused the drop
4. Tell me one thing I should watch closely next week to test the most likely hypothesis

What good output looks like: The AI confirms whether the drop is statistically meaningful relative to your prior weeks (not just how it felt), identifies which specific metric fell most, and offers hypotheses tied to what you described — not generic stress-and-sleep advice. If the output skips step 1 and jumps straight to solutions, push back: “First confirm whether the drop is significant based on the numbers, then generate hypotheses.”

Prompt 4: Testing Whether an Intervention Worked

Use when: You made a deliberate change to your focus setup four or more weeks ago and have enough before-and-after data to assess whether it had a measurable effect.

Four weeks ago I made a specific change to my focus setup: [describe the change].

Here are my session logs from the two weeks before the change and the two weeks after:

[paste before data]
[paste after data]

Please compare my three Focus Dashboard metrics — deep hours per day, session
completion rate, distraction count per hour — between the two periods.

Tell me:
1. Whether there is a meaningful difference in any of the three metrics
2. Whether the change appears to have had an effect, a mixed effect, or no effect
3. What, if anything, is ambiguous and needs a longer test period to resolve

What good output looks like: A clear before/after comparison for each of the three Focus Dashboard metrics, a verdict on whether the change had an effect or no effect, and an honest statement about what the two-week window cannot rule out. A failed response gives an encouraging “it seems like it helped!” without comparing the actual numbers, or declares success based on one strong week in the after period.

Prompt 5: Monthly Trend Review

Use when: You are at the end of a month and want to know whether your focus capacity is improving, holding steady, or slowly eroding — and what the trend suggests to prioritize next month.

Here are four weeks of focus session logs. I want to understand my trend, not
just my weekly averages.

[paste four weeks of logs]

Please:
1. Calculate my three Focus Dashboard metrics for each individual week
2. Identify whether each metric is trending up, down, or flat across the month
3. Flag any week that looks anomalous relative to the others and suggest what
   might explain it
4. Tell me whether my overall focus performance at the end of this month is
   stronger, weaker, or similar to the beginning — and what one thing the data
   suggests I should focus on next month

What good output looks like: Week-by-week metric tables (not averaged across the month), a clear directional label (improving/flat/declining) for each metric, a flag on any anomalous week with a plausible explanation, and a single priority for next month — not a list of five things. A response that gives you a single monthly average for each metric has collapsed the trend data you submitted and is not doing the job this prompt asks for.

Common Pitfalls

Pitfall: Logging sessions but omitting context variables. If your log has session length and quality rating but no meeting count or sleep note, the AI cannot distinguish an off day from a structural problem. Fix: add two context lines at the top of each weekly log — meetings per day (average) and sleep quality (1–3 scale). Without them, Prompt 2 will return correlations based solely on time of day.

Pitfall: Running Prompt 4 with only one week before and one week after. A single week is too noisy to draw conclusions from — one bad week can reflect a conference, illness, or an unusual deadline. Fix: wait for two full weeks of data on each side before running the intervention assessment. The prompt says four weeks ago for a reason.

Pitfall: Accepting a qualitative response when you asked for a quantitative one. If you ask the AI to “calculate my average deep hours per day” and it responds with “it looks like you had several strong sessions,” the prompt did not work. Fix: reply with “Please calculate the exact number from the data I provided.” AI tools will sometimes summarize impressionistically when the data is hard to parse — do not accept it.

Pitfall: Treating the AI’s hypothesis as a diagnosis. Prompt 3 asks for hypotheses about why a week went wrong. Those are starting points for investigation, not conclusions. Fix: after the AI generates two or three hypotheses, ask “What data from next week would confirm or rule out each one?” That turns a guess into a test.

These five prompts cover the analytical questions that matter: where you stand now, what conditions produce your best work, what caused a bad period, whether a change is working, and whether you are improving over time. Pick the one that fits your current situation and run it before your next planning session.

For the full framework — including how to structure your session log, which metrics to track first, and what to do with the output over time — see the Complete Guide to Focus Metrics and AI.

Related: Complete Guide to Focus Metrics and AI · How to Measure Focus with AI · Beyond Time Focus Metrics Walkthrough

Tags: AI prompts, focus analysis, session logging, weekly review, deep work prompts