Great mockups start with a clear brief. If you want usable product visuals in one session, you need structure, not random experimentation. This workflow is designed for marketing teams who need consistent product imagery without a long handoff chain.
The steps below work in any AI image generator, but they map especially well to nano bannana because the workflow emphasizes prompt structure, reference guidance, and repeatable results.
Step 1: Define the mockup brief
Write down the product name, materials, and usage context. Keep it short. A good brief names the subject and placement in one sentence.
Example brief:
- Product: matte black smartwatch
- Context: studio hero image for a landing page
- Mood: premium, clean, minimal
- Format: 4:5 and 16:9
If you are working on a campaign, add the channel. A social post needs a different crop than a landing page hero.
Step 2: Collect reference cues
References help you stay consistent. Collect two or three visual cues that define your look:
- Brand colors or gradients
- Preferred lighting (soft, high contrast, natural)
- Texture references (stone, metal, fabric)
- Composition cues (centered, angled, flat lay)
You can use real product photos, style boards, or simple mockups. The goal is to communicate the visual direction with minimal ambiguity.
Step 3: Build a base prompt
Use a template that stays consistent across runs:
[product] mockup, [context], [style], [lighting], [materials],
clean background, [composition], [aspect ratio]Example:
matte black smartwatch mockup, studio hero, premium product photography,
soft light, brushed metal accents, clean background, centered composition, 4:5Keep the prompt short enough to scan and long enough to describe the essentials.
Step 4: Generate a small set of variations
Start with four to six variations. Keep the prompt stable and only change one element at a time:
- Lighting (soft vs contrast)
- Background (neutral vs gradient)
- Angle (front vs 3/4)
This gives you comparable outputs and avoids the trap of random changes.
Step 5: Refine the best candidate
Pick the strongest option and refine with small, specific adjustments:
- "subtle shadow under product"
- "slight reflection on surface"
- "soft vignette"
If you change too many variables at once, you lose the thread of what is working.
Step 6: Add brand tokens
Brand tokens are short phrases that define your visual system. Add one or two to reinforce consistency:
- "warm neutral palette"
- "matte texture"
- "clean studio light"
These tokens can be reused across all mockups in a campaign, which makes the visuals feel cohesive.
Step 7: Export the sizes you need
Generate dedicated outputs for each placement. Do not rely on heavy cropping later. A hero image and a social tile need different composition and negative space.
If you are working in nano bannana, save the prompt with the placement name so you can reuse it later.
Step 8: Run a quick quality check
Before you ship, check these basics:
- The product is readable at small sizes.
- The background leaves space for copy.
- The lighting matches your brand style.
- The composition is consistent across variants.
If one item fails, adjust one variable and regenerate.
Common issues and fixes
Product looks flat
Add lighting direction and shadow detail.
Background feels noisy
Use "clean background" and specify a single color.
Brand colors drift
Repeat the exact palette name in the prompt and avoid competing color adjectives.
Edges look inconsistent
Add a constraint like "sharp edges" or "clean product outline".
Workflow tips for teams
- Keep a shared prompt library with one template per product category.
- Store the final prompt with the exported assets so you can re create later.
- Use a simple review rubric so approvals are consistent.
Consistency is easier when the workflow is shared, not just the output.
When to use image to image
If you already have a product photo or a rough mockup, image to image can speed up iteration. It helps preserve shape, layout, and material cues while you explore styling changes. Use it when:
- The product has a specific silhouette that must remain accurate.
- The material finish is a core part of the brand story.
- You need multiple angles that still feel like the same object.
Describe what should remain consistent so the model does not drift.
Example prompt variations by category
Consumer electronics
[device] mockup, studio hero, cool neutral palette, crisp reflections,
clean background, centered composition, 4:5 aspect ratioBeauty and skincare
[product] bottle mockup, soft diffused light, warm neutral palette,
minimal background, gentle shadow, 4:5 aspect ratioFood and beverage
[product] packaging mockup, natural light, subtle texture backdrop,
clean composition, space for headline, 16:9 aspect ratioThese examples keep the structure stable while letting the category details drive the look.
Organize outputs for reuse
Name exports with the product, placement, and prompt version. This makes it easy to find the winning asset and reuse the prompt later. A simple naming pattern like product-placement-v2 is enough to keep a library clean.
Lighting notes that improve realism
Lighting is often the difference between a flat mockup and a believable one. Add a simple lighting cue such as "soft studio light from left" or "diffused top light." Keep it consistent across a set so the assets feel related. If your brand uses high contrast photography, specify it early and keep the background simple to avoid noise.
When in doubt, choose softer lighting first, then increase contrast only if needed.
Result checklist
- The product is the clear focal point.
- The background supports the message, not the other way around.
- The mockup aligns with the brand visual system.
For more prompt ideas, see Prompt Recipes.
