Product Photos with Nano Banana: Ecommerce Hero Scene Swaps

Learn how to build a usable ecommerce hero with Nano Banana: keep the real product, swap the backdrop for a scene that sells. Built for people who want AI product images and Nano Banana ecommerce visuals without redrawing packaging from scratch. Plan about 20–25 minutes.

Learning goals

• Prep one clean product reference
• State in one prompt what to keep and what scene to use
• Reach listing-ready quality in 2–3 conversational turns

What to prepare

• A clear front or three-quarter product photo (simple background helps)
• A target scene in words (kitchen counter, bathroom vanity, linen still life, outdoor wood table)
• Access from nanobanana-ai.us into the Nano Banana workbench

English workbench entry:

https://app.nanobanana-ai.us/en/generate/image-tools/nano/

Step 1: Choose the right reference (3 minutes)

Half of a good product scene swap is the reference:

• Full product in frame; labels and bottle text readable when possible
• Avoid extreme perspective and heavy occlusion
• Even light, few messy shadows

If you only have a busy studio shot, pick the clearest hero angle—don’t dump conflicting references in one go.

Step 2: Write a clear scene-swap prompt (5 minutes)

After you upload the product image, say three things in plain language: keep the subject / new scene / light mood.

Template

Keep the product shape, label, and materials from the reference unchanged. Replace the background with [new scene]. Use [lighting description], product centered, ecommerce hero composition, realistic product photography.

Example: skincare bottle on a bathroom vanity

Keep the glass skincare bottle shape, pump, and label text from the reference unchanged. Swap the background to a light marble bathroom vanity with a folded white towel beside it. Soft window light, shallow depth of field, ecommerce hero, realistic product photography.

Example: food jar on a kitchen counter

Preserve the jar print and lid details. Change the scene to a bright kitchen wood counter with softly blurred clean cabinets behind. Morning side light, product centered, realistic still life suitable for an ecommerce hero.

Rule of thumb: lock the product first, then swap the scene. That is why Nano Banana product images stay steadier than pure text-to-image for catalog work.

Step 3: Generate v1 and run a QA pass (5 minutes)

Check only four things:

1. Does the product still look like the real SKU? (silhouette, color, label)
2. Does the scene match the category mood?
3. Is the subject large enough and centered?
4. Any weird warping or broken text?

Product drifted: return to Step 2 and strengthen “keep appearance / label unchanged.”
Scene wrong: rewrite only the scene clause—don’t restyle light and props in the same turn.

Step 4: Two refine rounds to listing-ready (8 minutes)

Change one thing per turn:

Keep the product completely unchanged; only soften the countertop reflections a little

Keep product and scene; remove the extra prop on the right

Slightly sharpen the product, soften the background further, keep the ecommerce hero feel

That rhythm is the practical core of AI product images for shops: reference locks the SKU; conversation fixes mood and clutter.

Step 5: Export and pre-listing checks (2 minutes)

Before download / upload:

• Brand and packaging text intact—no scrambled glyphs
• No obvious fake hands or fake type
• Crop still works for your marketplace hero ratio

For seasonal or colorway series, reuse the same product reference and swap only the scene sentence while keeping the light template—series consistency improves fast.

FAQ

Why does the product warp after a scene swap?

The instruction was too soft, or you changed too much at once. Stress shape, label, and materials, and edit in separate turns.

Can I skip the product photo and use text only?

Fine for mood boards. For ecommerce heroes, a real packaging reference is safer for conversion and brand accuracy.

How is this different from Nano Banana text-to-image?

Text-to-image invents from words. This tutorial is reference + product scene swap—closer to real catalog production.

Practice list

• Prep 1 clean product reference
• Generate 1 scene-swapped hero from the template
• Finish 2 single-focus refine turns
• Export and run the four pre-listing checks

Wrap-up

For Nano Banana ecommerce heroes, flashy prompts matter less than a simple loop: lock product → swap scene → refine in small turns. Master that, and you can ship reliable AI product images into everyday listing workflows.