What the AI found
“Your top three days for focused work by flow state scores consistently involved a 20-minute morning walk, leading to an average 15% increase in concentration over days without a morning walk.”
Before
Subjective daily notes, rarely reviewed
After
Quantified impact of morning walks on focus
The same system, three states — real screens, not a screenshot
- Tuesday: felt sharp after early gym. good flow on coding sprint.
- Wednesday: abit groggy after late night. coffee helped, but focus dipped after lunch.
- Thursday: very productive. solid flow state this morning. walked to train.
- Friday: distraction city. too many meetings. caffeine jitters.
Prompt
I've been tracking my daily habits and subjective flow state scores in a Google Sheet for the past four weeks. I want to understand what influences my focus. Here's a sample of my data: date, morning_walk_taken (yes/no), total_focused_work_hours, flow_state_score (1-10). Find any strong correlations, especially regarding morning walks and flow state scores. [pasted csv data]
I've been tracking my daily habits and subjective flow state scores in a Google Sheet for the past four weeks. I want to understand what influences my focus. Here's a sample of my data: date, morning_walk_taken (yes/no), total_focused_work_hours, flow_state_score (1-10). Find any strong correlations, especially regarding morning walks and flow state scores. [pasted csv data]
AI
Upon analysing your data, a clear pattern emerges: your top three days for focused work, as indicated by flow state scores (averaging 8.7), consistently involved a 20-minute morning walk. On average, days with a morning walk showed a 15% increase in your 'flow_state_score' compared to days without, moving from an average of 6.2 to 7.1.7.1
Average Flow Score w/ walk
6.2
Average Flow Score w/o walk
+15%
Flow Score Increase
Weekly cognitive function review
From scattered observations to a focused understanding of daily habits influencing cognitive performance.
A 38-year-old software engineer, Northern Europe
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 software engineer, much like many in demanding roles, found themselves jotting down thoughts on their cognitive state throughout the week. Snippets about "good focus today" or "foggy start" would end up in a digital note-taking app, alongside meeting reminders and grocery lists. The intention was to review them, but without a clear system or a prompt, these valuable observations rarely translated into actionable insights. The data was there, but it was noisy and unanalysed.
- Tuesday: felt sharp after early gym. good flow on coding sprint.
- Wednesday: abit groggy after late night. coffee helped, but focus dipped after lunch.
- Thursday: very productive. solid flow state this morning. walked to train.
- Friday: distraction city. too many meetings. caffeine jitters.
Working state
Core Course, doing its job
The first step in the "Ledger" layer of the 3-Layer Method was to centralise these daily notes and augment them with a relevant, easy-to-track metric. Using Google Sheets for its simplicity, the engineer established a daily logging rhythm, capturing "focused work hours" and a subjective "flow state score" alongside notes on key habits. The real breakthrough came when feeding a week's worth of this structured data into Gemini, prompting it to find patterns related to a specific hypothesis.
Prompt
I've been tracking my daily habits and subjective flow state scores in a Google Sheet for the past four weeks. I want to understand what influences my focus. Here's a sample of my data: date, morning_walk_taken (yes/no), total_focused_work_hours, flow_state_score (1-10). Find any strong correlations, especially regarding morning walks and flow state scores. [pasted csv data]
I've been tracking my daily habits and subjective flow state scores in a Google Sheet for the past four weeks. I want to understand what influences my focus. Here's a sample of my data: date, morning_walk_taken (yes/no), total_focused_work_hours, flow_state_score (1-10). Find any strong correlations, especially regarding morning walks and flow state scores. [pasted csv data]
AI
Upon analysing your data, a clear pattern emerges: your top three days for focused work, as indicated by flow state scores (averaging 8.7), consistently involved a 20-minute morning walk. On average, days with a morning walk showed a 15% increase in your 'flow_state_score' compared to days without, moving from an average of 6.2 to 7.1.Use case implemented
The finished system, running on its own
With Gemini having identified a clear pattern, the system moved into the "Protocol" phase. The engineer now reliably incorporates a 20-minute morning walk, knowing its quantifiable impact on their focused work. Each Sunday, a quick five-minute review in Google Sheets, followed by a brief check-in with Gemini, reinforces the habit and allows for iterative adjustments. The system now runs itself, providing consistent, evidence-based feedback on a key performance indicator.
7.1
Average Flow Score w/ walk
6.2
Average Flow Score w/o walk
+15%
Flow Score Increase
What an outside observer would notice
↑ 80%
Morning walk adherence
↑ 10%
Avg. flow state score
↓ 20%
Self-reported distraction
The stack — build it yourself
Accessible, flexible spreadsheet for structured daily logging.
Powerful AI for identifying non-obvious correlations in personal data.
Fast, local capture for initial, unstructured observations.
These are the tools used in this story. Any can be swapped for an equivalent you already trust.
Go deeper
Do this yourself
Explore the 3-Layer Method
This story runs on Core Course. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.