Lesson Content

Systematic prompt refinement process.

Effective refinement follows a process: Generate -> Review all four variations -> Identify the closest to intent -> Identify what specifically is wrong or missing -> Add, remove, or adjust specific prompt elements -> Regenerate -> Repeat. Changing too many elements at once makes it difficult to identify which change produced the improvement. Change one or two elements per iteration cycle for diagnostic refinement.

Identifying what to adjust.

Common adjustments based on results: Subject attribute issues (wrong age, expression, clothing, framing) – add more specific subject description; Style inconsistency – add or strengthen style language, or adjust Content Type setting; Lighting problems – add specific lighting description; Composition issues – add compositional language ("three-quarter portrait framing," "wide establishing shot," "close-up detail"); Mood mismatch – add explicit mood or atmosphere language.

Building a personal prompt library.

Effective prompt elements discovered through refinement are worth documenting for reuse: prompt patterns that consistently produce the visual style needed for a specific client or use case; effective lighting descriptions; style combinations that work well; subject description patterns for common subject types. A documented prompt library reduces refinement cycles for future similar projects.

Evaluating professional readiness.

Before using Firefly-generated images in professional work: visual quality appropriate for the specific use context (social media, website, print – each has different quality thresholds); subject and content appropriate for the brand and audience; no visible AI generation artifacts (unnaturally blended elements, incorrect anatomy, text rendering errors); and commercial use verified per current Adobe Firefly terms. Professional readiness is a judgment call – apply the same quality standard you'd apply to stock photography.