Name what good looks like
Outcome
Before you prompt, you can describe what a good answer looks like — which makes everything after it easier.
Concept
Most people open a chat and start typing a request. The skill that separates good AI work from frustrating AI work happens one step earlier: knowing what you actually want before you ask. If you can't describe what an excellent answer would contain, you can't ask for it clearly, you can't tell whether what comes back is any good, and you can't hand the task off and trust the result. This one habit is the foundation of every other skill here.
Weak prompt vs. strong prompt
The ask
Summarize this article.
What comes back
This article discusses the impact of remote work on urban economies, touching on various factors such as office vacancy rates, local business revenues, housing trends, and the broader implications for city planning. The author presents several perspectives and concludes that the shift is complex and ongoing.
Try it
Pick a task you'd hand to an AI. Before you write the prompt, write down what a good answer would contain — the specifics, the shape, what it must include, what it should leave out. Then ask, and hold what comes back against what you wrote.
Takeaway
Naming what good looks like, before you ask, is the move underneath every other skill. Do it first and everything downstream gets easier.
Your turn
This prompt makes Claude guess. Rewrite it so it has a definition of good to answer your actual situation — the checklist tracks what you add.
The weak prompt
Give me feedback on my essay.
What you've added
- Names the qualities of a good answerSay what makes a critique useful: specific, balanced, prioritized.
- Describes the shape a good answer takese.g. point to exact lines, or end with the changes that matter most.
- Says what would make it a bad answerName what to avoid: vague praise, a generic summary.
- Sets a bar or comparisonCompare to a standard: 'like a tough editor would.'