Master the art of writing prompts that generate better thumbnails
Effective prompts are the foundation of great thumbnails. This guide teaches you how to write prompts that consistently produce results you love.
Bad: "Add the logo somewhere visible" Good: "(Logo) positioned in top-right corner, 20% of image width"
When you reference images, be specific about placement and size. The AI needs clear instructions about where and how large elements should appear.
Bad: "Create a thumbnail that looks amazing with good composition and excellent lighting" Good: "Bright golden hour lighting from top-left. Balanced composition with person on left, product on right"
Remove subjective words like "amazing," "excellent," or "nice." Instead, describe exactly what you want.
Bad: "Make it pop" Good: "High contrast between person and background. Vibrant saturated colors"
"Pop" means different things to different people. Describe the actual visual qualities you want.
Bad: "Change the expression, the background color, the lighting, and add a new element" Good: Focus on one or two changes per iteration
Multiple changes at once make it hard to understand which change caused the result. Iterate incrementally.
Bad: "Add bold white text that is large and prominent and stands out" Good: "Bold white text reading '24 HOURS ONLY' at top, high contrast with background"
Be concise. One clear instruction per element is more effective than repeating the same idea multiple ways.
Always use parentheses for image names:
(Logo) in top-right corner
(Product Image 1) centered on table
(Icon Set) arranged along bottom edge
The parentheses tell the AI: "This is a reference to an uploaded image."
Describe what you want, not what you don't want:
Vague: "No dark colors" Specific: "Bright red, yellow, and white color palette"
Vague: "Make it energetic" Specific: "Dynamic diagonal composition. Sharp angles. High contrast colors"
When iterating, isolate changes:
Before: "The background is blue, the person has a happy expression, and there's a floating element" Better for editing: "Replace the blue background with a red gradient. Keep the person's happy expression. Keep the floating element."
This approach helps the AI understand exactly what's changing.
If you refer to elements multiple times, use the same name:
Inconsistent: "Add the logo in the corner, make sure the brand image is visible" Consistent: "Add (Logo) in top-right corner. (Logo) should be clearly visible"
Here's a real-world example of iterating from concept to final result:
Person with excited expression looking at (Product Image).
Bold text "BEST DEAL" at top.
Simple background.
Result: Works well but feels flat.
Changes made: Add lighting, color intensity
Person with excited expression looking at (Product Image).
Bold text "BEST DEAL" at top in white.
Vibrant red background with diagonal geometric shapes.
Golden lighting from top-left highlighting the person's face.
Result: Much better! But text needs more emphasis.
Changes made: Enhance text, refine composition
Person with excited expression. Golden lighting from top-left.
Bold text "BEST DEAL" at top - white with shadow, stands out against background.
(Product Image) centered, mid-size.
Vibrant red background with diagonal geometric shapes.
Result: Strong result. One final refinement.
Changes made: Fine-tune expression and composition
Person with shocked, delighted expression. Golden lighting from top-left.
"BEST DEAL" text at top - large, white with black shadow for contrast.
(Product Image) positioned right of center, clearly visible.
Vibrant red background with diagonal geometric lines creating movement.
Result: Final thumbnail ready for use!
Person with shocked, wide-eyed expression.
Warm golden lighting from top-left side.
"SHOCKING TRUTH" text at top in bold white with black outline.
(Evidence Image) positioned center-right, clearly visible.
Deep blue background with lightning bolt elements scattered throughout.
High contrast. Dramatic, urgent composition.