Style Principles
Apply these guidelines when generating business visuals regardless of domain.
Professional Aesthetic
- Clean and modern — Always specify "clean, modern, professional" to avoid cluttered results
- Flat illustration preferred — Flat/illustrated styles produce more consistent results than photorealistic
- Neutral backgrounds — Specify "white background" or "light gray background" for versatility
- Minimize text — AI text rendering is imperfect; keep text labels short and simple
Composition
- Specify orientation — Landscape for presentations, square for social/Slack, portrait for documents
- Leave whitespace — Don't overcrowd the composition; breathing room improves clarity
- Hierarchy matters — Main subject should be clear; supporting elements should recede
- Consider context — The visual will appear in slides, wikis, or printed materials—design accordingly
Color and Branding
- Default to safe palettes — Blues, teals, and grays read as professional and trustworthy
- Specify brand colors — If user provides brand colors, incorporate them
- Limit palette — 2-3 main colors plus neutrals; avoid rainbow effects
- Consider accessibility — Ensure sufficient contrast between elements
Technical Considerations
- Output quality — Request high resolution if known print/display context
- Consistency across series — When generating multiple related visuals, maintain style continuity
- Acknowledge limitations — AI-generated images may need manual touch-up for text or precise data
Visual Types That Work Well
- Diagrams: Funnels, flowcharts, process flows, org charts
- Illustrations: Personas, conceptual art, abstract representations
- Maps: Territory coverage, regional breakdowns (stylized, not cartographically accurate)
- Infographics: Timeline-based, comparison grids, simple data visualization
- Badges/Awards: Recognition graphics, achievement icons
Visual Types That Struggle
- Precise data charts — Use charting tools instead; AI struggles with exact numbers
- Detailed text — Anything requiring readable paragraphs
- Photo editing — Precise manipulation of existing photos
- Brand logos — Won't render existing logos accurately