MODULE 01LESSON 02 · 12:40

Effective Prompting

The 4-part framework beginners use to get great output on the first try.

SYSTEM_PLAYBACK_ACTIVEVideo placeholder

Prompting Lovable is a design skill, not a coding skill. You're describing an outcome — who it's for, what it should do, and what to leave alone. The more clearly you frame the ask, the less back-and-forth you need.

Beginners typically make one of two mistakes: prompts that are too vague ("make it nicer") or too crowded (a wall of ten requirements at once). Both slow you down. The fix is a simple structure you can use for every request.

/The C.O.C.E. framework

  • 01Context: Who is the user and what are they trying to accomplish on this screen?
  • 02Objective: The single measurable outcome for this prompt — one thing, not five.
  • 03Constraints: What must NOT change — colors, routes, existing components, tone.
  • 04Examples: Reference apps, screenshots, or a URL for Lovable to model against.

/A weak prompt vs. a strong prompt

text
Weak:   "Make a nice landing page for my app."

Strong: "Build a landing page for a habit tracker aimed at students.
         Hero, three feature cards, testimonial, CTA to /signup.
         Keep the existing color tokens and typography.
         Reference vibe: linear.app — calm, high contrast, lots of whitespace."
Beginner tip

One request per prompt. Small, focused edits produce far better results than a wall of requirements. If you catch yourself writing "and also…", stop and split it into a second prompt.

/Iterate, don't restart

When something isn't quite right, describe the delta — "the hero is too tall, make it 60vh and left-align the headline" — rather than re-prompting from scratch. Lovable already has the context; you just need to nudge it.