Building the habit.
DAY ONE
One real task. Something you do regularly that involves writing.
Notice how much the output improves when you add context.
DAYS TWO TO THREE
The same task, done the same way each time.
Start noticing which phrasings produce better results.
DAYS FOUR TO FIVE
Try a second kind of task: summarising a document, or explaining something unfamiliar.
END OF THE WEEK
Save the prompt that worked best. You will reuse it.
WHAT TO AVOID IN THE FIRST WEEK
Trying everything at once Automating anything Publishing output you have not read
WHAT TO EXPECT
Useful results immediately. Noticeably better results as your prompting improves.
WHAT TO MEASURE
How long the task took before, and how long it takes now including checking.
THE HABIT THAT MATTERS MOST
Providing context. Most disappointing output comes from prompts that assumed knowledge the model did not have.
