What the AI found
“Your sleep quality dips significantly after nights when you read fiction past 10 PM, not from screen light, but from the emotional engagement delaying the wind-down process.”
Before
Randomly tired, blaming usual suspects
After
Targeted nightly routine, consistently rested
The same system, three states — real screens, not a screenshot
| Wake Time | 07:15 |
| Bed Time | 23:30 |
| Sleep Quality (1-5) | 3 |
| Last Meal Time | 19:00 |
Prompt
Analyse my sleep log. For each night, I have: Wake Time, Bed Time, Sleep Quality (1-5), Last Meal Time, and Evening Activity. Identify any unexpected correlations between Evening Activity and Sleep Quality. Here's the data: Date,Wake Time,Bed Time,Sleep Quality,Last Meal Time,Evening Activity 10/01,07:00,23:00,4,19:30,Light reading 10/02,07:30,00:00,2,20:00,Thriller novel 10/03,07:15,22:45,5,19:00,Meditating 10/04,07:45,00:30,2,20:30,Action movie 10/05,07:00,23:00,4,19:00,Documentary 10/06,08:00,00:15,3,20:00,Strategy game 10/07,07:00,22:30,5,18:30,Calm music 10/08,07:15,23:15,3,19:00,Fantasy novel 10/09,07:30,00:00,2,20:00,Intense debate 10/10,07:00,22:45,4,19:30,Podcast
Analyse my sleep log. Identify any unexpected correlations between Evening Activity and Sleep Quality.
AI
I noticed a consistent pattern: on nights with 'Thriller novel' or 'Fantasy novel' as your evening activity past 10 PM, your average sleep quality rating dropped to 2.3. This is significantly lower than your average of 4.2 on nights with 'Light reading' or 'Calm music', suggesting the content's engaging nature, not just screen time, is impacting your sleep onset and quality.4.5 (↑1.2)
Average Sleep Quality
5/7 (↑2)
Nights with Deep Sleep > 90 min
18 min (↓15 min)
Time to Sleep Onset
Optimising Sleep: How a Playbook Uncovered a Hidden Habit
A pre-built AI playbook transformed guesswork into precision sleep improvement.
A 38-year-old marketing director, Northern Europe, keen on improving sleep without drastic lifestyle changes.
Tools used
The real tools used here — swap any for your own equivalent. Each links to how we’d set it up.
Starting state
Before anything was set up
Our subject, a marketing director, often felt fatigued despite trying various sleep hygiene tips. They diligently avoided screens, caffeine, and heavy meals before bed, yet sleep remained elusive. Their sleep tracker showed inconsistency, but no clear patterns emerged, leading to frustration and a sense of resigned acceptance that "this is just how I sleep."
| Wake Time | 07:15 |
| Bed Time | 23:30 |
| Sleep Quality (1-5) | 3 |
| Last Meal Time | 19:00 |
| Evening Activity | TV, reading |
Working state
Resources, doing its job
The Resources playbook was deployed, prompting a structured data entry into Google Sheets. Instead of vague notions, specific daily activities and their timing were logged. The playbook then directed analysis via ChatGPT, where the data was cross-referenced with sleep tracker outputs. The AI, with its analytical speed, quickly sifted through weeks of data, looking for non-obvious correlations.
Prompt
Analyse my sleep log. For each night, I have: Wake Time, Bed Time, Sleep Quality (1-5), Last Meal Time, and Evening Activity. Identify any unexpected correlations between Evening Activity and Sleep Quality. Here's the data: Date,Wake Time,Bed Time,Sleep Quality,Last Meal Time,Evening Activity 10/01,07:00,23:00,4,19:30,Light reading 10/02,07:30,00:00,2,20:00,Thriller novel 10/03,07:15,22:45,5,19:00,Meditating 10/04,07:45,00:30,2,20:30,Action movie 10/05,07:00,23:00,4,19:00,Documentary 10/06,08:00,00:15,3,20:00,Strategy game 10/07,07:00,22:30,5,18:30,Calm music 10/08,07:15,23:15,3,19:00,Fantasy novel 10/09,07:30,00:00,2,20:00,Intense debate 10/10,07:00,22:45,4,19:30,Podcast
Analyse my sleep log. Identify any unexpected correlations between Evening Activity and Sleep Quality.
AI
I noticed a consistent pattern: on nights with 'Thriller novel' or 'Fantasy novel' as your evening activity past 10 PM, your average sleep quality rating dropped to 2.3. This is significantly lower than your average of 4.2 on nights with 'Light reading' or 'Calm music', suggesting the content's engaging nature, not just screen time, is impacting your sleep onset and quality.Use case implemented
The finished system, running on its own
With the playbook fully implemented, the subject now has a precise understanding of their sleep triggers. The nightly routine includes an earlier cutoff for mentally engaging activities, replaced by passive wind-down. The initial guesswork is gone, replaced by data-driven adjustments that yield consistent, higher-quality sleep, monitored through their sleep tracker and regular check-ins with the playbook.
4.5 (↑1.2)
Average Sleep Quality
5/7 (↑2)
Nights with Deep Sleep > 90 min
18 min (↓15 min)
Time to Sleep Onset
What an outside observer would notice
Increased by 1.2 points (on 5-point scale)
Average Sleep Quality
Improved from 3 to 5 per week
Number of good sleep nights (score 4+)
Consistently higher throughout the day
Subjective feeling of alertness
The stack — build it yourself
Provided the structured framework and analytical prompts for sleep improvement.
Easy to set up custom logging for specific daily activities and sleep metrics.
Quickly sifted through logged data to find non-obvious correlations affecting sleep.
Provided reliable, passive sleep data to validate subjective feelings and AI findings.
These are the tools used in this story. Any can be swapped for an equivalent you already trust.
Go deeper
Do this yourself
See Resources
This story runs on Resources. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.