Product Marketing Master Guide

The definitive guide to AI-powered product marketing image generation with Go Bananas!

This comprehensive guide covers everything you need to know about creating consistent, professional product marketing images using the Product Reference System. Whether you're an e-commerce store owner, marketing professional, or AI automation developer, this guide will help you master product image generation.

Part 1: Foundations
Part 2: Quick Start
Part 3: Source Image Quality
Part 4: Creating Product References
Part 5: Generating Marketing Images
Part 6: Product Photography Scenarios
Part 7: Flash vs Pro for Products
Part 8: Advanced Techniques
Part 9: Platform-Specific Guidelines
Part 10: Troubleshooting Guide
Part 11: Tools Reference
Part 12: Use Case Gallery
Appendix A: Parameter Reference
Appendix B: Prompt Templates Library
Appendix C: Troubleshooting Decision Tree
Appendix D: Glossary

Part 1: Foundations

What is the Product Reference System?

The Product Reference System is a powerful feature of Go Bananas! that enables you to:

Save product images permanently - Store product photos from URLs in R2 cloud storage
Generate unlimited marketing scenes - Create consistent product imagery across any setting
Maintain visual fidelity - Preserve exact product colors, materials, and design
Build a product library - Organize products with names, descriptions, and tags

Why Product Marketing Matters

Product marketing images are critical for:

Use Case	Impact
E-commerce listings	Higher conversion rates with quality imagery
Social media content	Increased engagement with lifestyle shots
Advertising campaigns	Consistent branding across platforms
Email marketing	Better click-through rates
Print materials	Professional catalog and brochure content

The Problem Without Product References

WITHOUT Product References:
┌────────────────────────────────────────────────────────────────┐
│ Upload product → Generate scene → Re-upload → Generate again   │
│ • Each upload is temporary                                      │
│ • Results may vary between sessions                            │
│ • No consistency tracking                                       │
│ • Manual re-uploading required                                  │
└────────────────────────────────────────────────────────────────┘

WITH Product References:
┌────────────────────────────────────────────────────────────────┐
│ Save product ONCE → Generate UNLIMITED marketing scenes        │
│ • Permanent storage in R2                                       │
│ • Consistent representation every time                         │
│ • Usage tracking and analytics                                  │
│ • Works across all sessions                                     │
└────────────────────────────────────────────────────────────────┘

How It Works (Architecture Overview)

┌─────────────────────────────────────────────────────────────────┐
│                    PRODUCT REFERENCE FLOW                        │
├─────────────────────────────────────────────────────────────────┤
│                                                                  │
│  1. CREATE PRODUCT REFERENCE                                     │
│     ┌──────────────┐    ┌──────────────┐    ┌──────────────┐   │
│     │ Product URL  │ →  │   Fetch &    │ →  │  Store in    │   │
│     │ (External)   │    │   Validate   │    │  R2 + D1     │   │
│     └──────────────┘    └──────────────┘    └──────────────┘   │
│                                                                  │
│  2. GENERATE WITH PRODUCT                                        │
│     ┌──────────────┐    ┌──────────────┐    ┌──────────────┐   │
│     │ Scene Prompt │ →  │ Load Product │ →  │   Gemini     │   │
│     │ + Product ID │    │ from R2      │    │   Generate   │   │
│     └──────────────┘    └──────────────┘    └──────────────┘   │
│                                                                  │
│  3. RESULT                                                       │
│     ┌──────────────────────────────────────────────────────┐    │
│     │ Marketing image with exact product representation     │    │
│     │ • Consistent colors, materials, proportions          │    │
│     │ • Stored in R2 with full metadata                    │    │
│     │ • Linked to product reference for analytics          │    │
│     └──────────────────────────────────────────────────────┘    │
│                                                                  │
└─────────────────────────────────────────────────────────────────┘

Storage Architecture:

Component	Storage	Purpose
Full-resolution image	R2 Bucket	Source for AI generation
Thumbnail	R2 Bucket	Quick previews in UI
Metadata	D1 Database	Search, organization, analytics
Usage logs	D1 Database	Track generation count

Key Concepts: Products vs Characters

Feature	Products	Characters
Source	URL download (external photo)	AI generation
Consistency	Exact visual match	Approximate likeness
Reference images	Single product image	Multiple reference images
Best for	Physical products, merchandise	People, creatures, mascots
Modifications	None (exact product preserved)	Scene adaptations allowed
Use case	E-commerce, marketing	Storytelling, branding

Part 2: Quick Start

3-Step Workflow

STEP 1: Save Product
─────────────────────
"Save product 'Premium Headphones' from https://example.com/headphones.jpg"

STEP 2: Generate Marketing Image
─────────────────────────────────
"Generate Premium Headphones on a minimalist desk with soft lighting"

STEP 3: Iterate & Scale
─────────────────────────
"Generate Premium Headphones in a gym setting"
"Generate Premium Headphones floating on gradient background"
"Generate Premium Headphones being worn at a coffee shop"

Your First Product

Step 1: Find a product image URL

Your product image should be:

Publicly accessible (direct URL to image file)
High resolution (1000px+ recommended)
Clean background (white or transparent ideal)
Well-lit with accurate colors

Step 2: Create the product reference

Using MCP (Claude Desktop, etc.):

Save this product image: https://example.com/water-bottle.jpg
Name: Premium Water Bottle
Description: Stainless steel 32oz bottle with bamboo cap
Tags: drinkware, outdoor, eco-friendly

Using REST API:

bash

curl -X POST "https://gobananasai.com/api/products" \
  -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "product_name": "Premium Water Bottle",
    "product_url": "https://example.com/water-bottle.jpg",
    "product_description": "Stainless steel 32oz bottle with bamboo cap",
    "tags": "drinkware,outdoor,eco-friendly"
  }'

Step 3: Verify the product was saved

Show me all my product references

You should see:

Product References (1 total):

1. **Premium Water Bottle** (ID: 1)
   Used: 0 times | Tags: drinkware, outdoor, eco-friendly
   "Stainless steel 32oz bottle with bamboo cap"

Your First Marketing Image

Generate a simple scene:

Generate the Premium Water Bottle on a clean white background

What happens internally:

System loads product reference from D1 database
Downloads product image from R2 storage
Builds specialized prompt with fidelity instructions
Sends to Gemini with reference image
Returns marketing image with exact product representation

Generated prompt (internal):

"Using the provided reference image showing Premium Water Bottle,
generate the product on a clean white background.
Maintain the exact design, colors, materials, textures,
and proportions of the product."

Behind the Scenes

When you save a product:

1. FETCH: Download image from provided URL
2. VALIDATE: Check it's a valid image (PNG, JPG, WebP)
3. PROCESS: Extract dimensions, generate thumbnail
4. STORE: Upload full image + thumbnail to R2
5. RECORD: Save metadata to D1 database

When you generate with a product:

1. LOOKUP: Find product by name or ID in D1
2. RETRIEVE: Download reference image from R2
3. BUILD: Construct prompt with fidelity instructions
4. GENERATE: Send to Gemini with reference image
5. SAVE: Store result in R2, link to product
6. TRACK: Update product usage statistics

Part 3: Source Image Quality

Ideal Product Images

The quality of your source image directly affects generation results. Here's what makes an ideal product image:

Resolution Requirements

Quality Level	Minimum Resolution	Recommended	Use Case
Basic	500 x 500	800 x 800	Quick prototypes
Standard	800 x 800	1200 x 1200	Social media, web
Professional	1200 x 1200	2000 x 2000	Print, advertising
Premium	2000 x 2000	4000 x 4000	Large format, catalogs

Rule of thumb: Use at least 1000px on the shortest side for professional results.

Background Best Practices

EXCELLENT (Recommended):
┌─────────────────────────────────────────┐
│ ✅ Pure white background (#FFFFFF)       │
│ ✅ Transparent background (PNG)          │
│ ✅ Light gray (#F5F5F5 - #EEEEEE)        │
│ ✅ Solid neutral color                   │
└─────────────────────────────────────────┘

ACCEPTABLE:
┌─────────────────────────────────────────┐
│ ⚠️ Simple gradient                       │
│ ⚠️ Subtle texture (marble, wood)        │
│ ⚠️ Soft drop shadow                     │
└─────────────────────────────────────────┘

AVOID:
┌─────────────────────────────────────────┐
│ ❌ Cluttered/busy backgrounds            │
│ ❌ Other products in frame               │
│ ❌ Text overlays or watermarks           │
│ ❌ Strong shadows or reflections         │
│ ❌ Lifestyle scenes (use plain shots)    │
└─────────────────────────────────────────┘

Lighting Guidelines

Good Lighting:

Even, diffused lighting from multiple angles
Soft shadows that don't obscure details
True color representation (no color cast)
Visible texture and material qualities

Bad Lighting:

Harsh directional lighting with deep shadows
Overexposure (blown highlights)
Underexposure (lost details in shadows)
Color cast from colored light sources

Product Positioning

IDEAL POSITIONING:
┌─────────────────────────────────────────┐
│                                         │
│           ┌───────────┐                │
│           │           │                │
│           │  PRODUCT  │                │
│           │  CENTERED │                │
│           │           │                │
│           └───────────┘                │
│                                         │
│  • Product fills 60-80% of frame       │
│  • Equal margins on all sides          │
│  • Primary angle showing key features  │
│                                         │
└─────────────────────────────────────────┘

Common Image Problems

Problem	Symptom	Solution
Low resolution	Blurry/pixelated output	Use higher resolution source
Color cast	Wrong colors in generations	Re-shoot with neutral lighting
Busy background	Product hard to isolate	Use white/transparent background
Harsh shadows	Unnatural lighting	Use diffused lighting setup
Cropped product	Missing parts in output	Show complete product in frame
Multiple products	Wrong product selected	One product per reference image
Watermarks	Text appears in output	Use clean, unmarked images
Heavy filters	Inconsistent appearance	Use unfiltered original photos

Source Image Checklist

Before saving a product reference, verify:

□ Resolution is at least 1000 x 1000 pixels
□ Background is clean (white, transparent, or solid neutral)
□ Product is centered and fills 60-80% of frame
□ Lighting is even without harsh shadows
□ Colors are accurate (no color cast)
□ Product is complete (nothing cropped)
□ No watermarks, logos, or text overlays
□ No other products or objects in frame
□ Image is in focus with visible details
□ File format is PNG, JPG, or WebP

Image Format Comparison

Format	Best For	Pros	Cons
PNG	Products on white/transparent	Lossless, transparency support	Larger file size
JPG	Lifestyle shots, photos	Smaller file size	No transparency, lossy
WebP	Web-optimized content	Best compression, transparency	Less universal support

Part 4: Creating Product References

From URL (Step-by-Step)

Method 1: MCP Tools (Conversational)

Basic creation:

Save product "Blue Gaming Mouse" from https://example.com/mouse.jpg

With description:

Create a product reference for "Ceramic Coffee Mug" from
https://example.com/mug.png with description "White ceramic mug
with minimalist design, 12oz capacity"

With tags:

Save product reference:
- Name: "Wooden Phone Stand"
- URL: https://example.com/phone-stand.jpg
- Description: "Handcrafted walnut wood phone stand with cable slot"
- Tags: "accessories,wood,handmade,desk-organization"

Method 2: REST API

bash

curl -X POST "https://gobananasai.com/api/products" \
  -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "product_name": "Wireless Speaker",
    "product_url": "https://example.com/speaker.png",
    "product_description": "Portable Bluetooth speaker with 360-degree sound",
    "tags": "audio,electronics,portable,bluetooth"
  }'

Response:

json

{
  "data": {
    "id": 16,
    "productName": "Wireless Speaker",
    "productUrl": "https://example.com/speaker.png",
    "r2Key": "acme-corp/2024-01-15/product-e5f6g7h8.png",
    "r2ThumbnailKey": "acme-corp/2024-01-15/product-e5f6g7h8-thumb.jpg",
    "width": 1000,
    "height": 1000,
    "sizeBytes": 412672,
    "thumbnailSizeBytes": 65536,
    "format": "png",
    "message": "Product reference created successfully"
  }
}

Product Naming Conventions

Good naming practices:

✅ GOOD NAMES:
─────────────────────────────────────────────
"Chrome Arrow Wall Decor Large"     - Descriptive, specific
"Blue Wireless Gaming Mouse RGB"    - Includes key features
"Handcrafted Walnut Phone Stand"    - Material and style
"Premium Leather Wallet - Brown"    - Color variant specified
"EcoTech Smart Water Bottle 32oz"   - Brand, type, size

❌ BAD NAMES:
─────────────────────────────────────────────
"Product 1"         - Too generic
"Arrow"             - Not descriptive enough
"Mouse"             - Could be anything
"test"              - Non-descriptive
"IMG_1234"          - Filename, not product name

Naming formula:

[Brand/Material] + [Product Type] + [Key Feature] + [Variant]

Examples:
- "Bamboo" + "Cutting Board" + "Large" = "Bamboo Cutting Board Large"
- "Wireless" + "Earbuds" + "Pro" + "Black" = "Wireless Earbuds Pro Black"

Writing Effective Descriptions

Descriptions help the AI understand your product for better generations.

Description template:

[Material/Construction] + [Key Visual Features] + [Size/Dimensions] + [Notable Details]

Examples:

BASIC (Acceptable):
"A water bottle"

GOOD:
"Stainless steel water bottle with matte black finish"

EXCELLENT:
"Stainless steel insulated water bottle, matte black finish
with rose gold accents, 24oz capacity, vacuum-sealed cap with
bamboo top, subtle embossed logo on side"

What to include:

Material (steel, ceramic, leather, wood, plastic)
Finish (matte, glossy, brushed, textured)
Colors (primary color, accent colors)
Size/dimensions if relevant
Notable design features
Brand elements (if important)

Tagging Strategy

Tags enable organization and searchability.

Tag categories:

Category	Examples
Product type	drinkware, electronics, apparel, furniture
Material	leather, wood, ceramic, metal, fabric
Use case	fitness, office, outdoor, travel, home
Campaign	summer-2024, holiday-sale, new-arrivals
Status	active, archived, seasonal, limited-edition
Brand	brand-name, collection-name

Tagging best practices:

✅ DO:
- Use lowercase, hyphenated tags: "eco-friendly"
- Be consistent across products: always "drinkware" not sometimes "drinks"
- Include multiple relevant categories
- Tag by use case for easy filtering

❌ DON'T:
- Use spaces: "eco friendly" (use "eco-friendly")
- Create one-off tags that won't be reused
- Over-tag with 20+ tags per product
- Use product-specific tags like "blue-mouse-1"

Multiple Product Variants

For products with variants (colors, sizes), create separate references:

Product: Water Bottle
├── "Water Bottle - Matte Black"
├── "Water Bottle - Rose Gold"
├── "Water Bottle - Ocean Blue"
└── "Water Bottle - Forest Green"

Variant naming convention:

"[Base Product Name] - [Variant]"

Examples:
- "Premium Headphones - Midnight Black"
- "Premium Headphones - Pearl White"
- "Premium Headphones - Space Gray"

Updating Products

Update metadata (name, description, tags):

Via REST API:

bash

curl -X PUT "https://gobananasai.com/api/products/16" \
  -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "product_description": "Updated description with more details",
    "tags": "audio, electronics, premium, portable"
  }'

Replace product image:

To update the actual product image, you must:

Delete the existing product reference
Create a new one with the updated URL

Delete product reference "Old Product Name"
Save product "Product Name" from https://example.com/new-image.jpg

Part 5: Generating Marketing Images

Basic Generation

Simple scene:

Generate the Premium Water Bottle on a clean white background

Lifestyle shot:

Show the Premium Water Bottle on a hiking trail

Product in context:

Generate the Premium Water Bottle at a gym with workout equipment

Scene Prompts That Work

Scene prompt formula:

[Product Name] + [Location/Setting] + [Context Elements] + [Atmosphere/Lighting]

Examples by complexity:

SIMPLE:
"Generate the Blue Gaming Mouse on a desk"

GOOD:
"Generate the Blue Gaming Mouse on a clean gaming desk
with RGB lighting"

EXCELLENT:
"Generate the Blue Gaming Mouse on a premium gaming desk setup
with RGB lighting, mechanical keyboard visible, and ambient
purple/blue glow in a modern home office setting"

Scene elements to include:

Element	Examples
Surface	wooden table, marble counter, glass shelf
Context	with laptop, near coffee cup, among books
Environment	home office, gym, outdoor cafe
Lighting	soft window light, dramatic studio, warm ambient
Atmosphere	cozy, professional, energetic, minimalist
Time	morning light, golden hour, evening ambiance

Additional Details

The additional_details parameter adds styling without changing the scene:

Generate the Premium Water Bottle on a rustic wooden table

Additional details: "warm golden hour lighting, shallow depth of field,
professional product photography, lifestyle magazine style"

Useful additional details:

Category	Examples
Photography style	"professional product photography", "lifestyle magazine style", "high-end advertising"
Lighting style	"soft studio lighting", "dramatic shadows", "natural window light"
Focus/DOF	"shallow depth of field", "product in sharp focus", "bokeh background"
Mood	"warm atmosphere", "clean minimalist", "cozy feeling"
Quality	"high detail", "premium aesthetic", "commercial quality"

Aspect Ratios by Platform

Platform	Aspect Ratio	Use Case
Instagram Post	`1:1` (square)	Feed posts
Instagram Story	`9:16` (portrait)	Stories, Reels
Pinterest	`2:3` or `9:16`	Pins
Facebook Feed	`1:1` or `4:5`	Posts
Facebook Ad	`16:9`	Video thumbnails, ads
Twitter/X	`16:9`	Posts with images
LinkedIn	`1:1` or `16:9`	Posts
Amazon	`1:1`	Product listings
Shopify	`1:1`	Product pages
Email Header	`16:9`	Newsletter headers
Print Catalog	`4:3`	Catalogs, brochures

Example:

Generate the Premium Headphones in a minimalist studio setup,
aspect_ratio: "1:1"

Negative Prompts

Negative prompts tell the AI what to avoid:

Generate the product on a desk,
negative_prompt: "blurry, distorted, unrealistic proportions,
wrong colors, damaged, low quality"

Useful negative prompts:

Issue to Avoid	Negative Prompt
Quality issues	"blurry, low quality, pixelated, amateur"
Product distortion	"distorted, wrong proportions, warped, stretched"
Color problems	"wrong colors, color shift, oversaturated, washed out"
Distracting elements	"cluttered, busy background, competing focal points"
Unwanted additions	"text, watermarks, logos, people unless specified"
Style conflicts	"cartoon, anime, illustration" (for photo-realistic)

Style Preset Integration

Combine products with style presets for brand consistency:

Generate the Premium Water Bottle on a minimalist desk,
style_preset_name: "Brand Product Photography"

Create a product-focused style preset:

Create style preset:
- Name: "Product Hero Shot"
- Prompt: "professional product photography, studio lighting,
  clean background, high detail, commercial quality"
- Negative prompt: "amateur, low quality, cluttered, distracting"
- System instruction: "Create premium product marketing images
  suitable for e-commerce and advertising"
- Aspect ratio: "1:1"

Batch Generation

Generate multiple scenes for a campaign:

Via MCP (sequential):

Generate the Premium Water Bottle on white studio background
Generate the Premium Water Bottle in a gym setting
Generate the Premium Water Bottle on a hiking trail
Generate the Premium Water Bottle at a beach picnic

Via REST API (batch):

bash

curl -X POST "https://gobananasai.com/api/products/batch-generate" \
  -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "product_ids": [1, 2, 3],
    "scene_prompt": "floating against a clean white background",
    "aspect_ratio": "square"
  }'

Via single product with multiple images:

Generate 4 variations of the Premium Water Bottle on a kitchen counter

Part 6: Product Photography Scenarios

E-commerce Hero Shots

Hero shots are the primary product images for listings:

STUDIO WHITE BACKGROUND:
"Generate the [Product] on a pure white background with
professional studio lighting and soft shadows"

FLOATING/LEVITATION:
"Generate the [Product] floating against a clean white
background with subtle shadow below"

GRADIENT BACKGROUND:
"Generate the [Product] on a soft gradient background
from white to light gray, professional product photography"

Lifestyle Photography

Show products in real-world use contexts:

HOME SETTING:
"Generate the [Product] on a modern living room coffee table
with plants and books, warm afternoon light"

OUTDOOR SETTING:
"Show the [Product] at a beach picnic with sand, ocean
backdrop, and summer vibes"

WORK SETTING:
"Generate the [Product] on a professional office desk
with laptop and notepad, natural window light"

Detail/Close-up Shots

Highlight product features:

TEXTURE FOCUS:
"Generate close-up of the [Product] showing the material
texture and finish details with dramatic side lighting"

FEATURE HIGHLIGHT:
"Create a detail shot of the [Product] focusing on the
[specific feature], macro photography style"

QUALITY SHOWCASE:
"Show the [Product] craftsmanship details with premium
lighting that highlights the quality"

Size Comparison Images

Help customers understand scale:

WITH COMMON OBJECT:
"Generate the [Product] next to a standard credit card
for size reference on white background"

IN HAND:
"Show the [Product] being held in a hand to demonstrate
size and grip"

CONTEXT SIZE:
"Generate the [Product] next to a smartphone to show
relative size comparison"

Environment Variations

Same product, different environments:

SEASONAL VARIATIONS:
Spring: "[Product] surrounded by fresh flowers and pastel colors"
Summer: "[Product] on a beach towel with sunglasses and waves"
Fall: "[Product] with autumn leaves and warm orange tones"
Winter: "[Product] on a cozy setting with knit blanket and snow outside"

TIME OF DAY:
Morning: "[Product] with breakfast setup and soft morning light"
Afternoon: "[Product] in bright natural sunlight"
Evening: "[Product] in warm ambient evening lighting"
Night: "[Product] in atmospheric low-light setting"

Seasonal Campaigns

HOLIDAY CAMPAIGN:
"Generate the [Product] in elegant gift box with ribbon,
festive holiday background with pine and ornaments"

SUMMER SALE:
"Show the [Product] in a vibrant summer scene with tropical
colors and beach elements"

BACK TO SCHOOL:
"Generate the [Product] on a student desk with notebooks,
pencils, and backpack"

BLACK FRIDAY:
"Create the [Product] on a dramatic dark background with
gold accents and premium lighting"

Platform-optimized imagery:

INSTAGRAM GRID (1:1):
"Generate the [Product] in aesthetic flat lay arrangement
with complementary accessories, top-down view, square format"

INSTAGRAM STORY (9:16):
"Show the [Product] in vertical lifestyle shot with person
using it, story-optimized composition"

PINTEREST (2:3):
"Create an inspiring Pinterest-style image of [Product]
in a beautiful lifestyle setting, vertical format"

FACEBOOK AD (16:9):
"Generate the [Product] hero shot with clean background
and space for text overlay, wide format"

Print Advertising

High-quality images for physical materials:

MAGAZINE AD:
"Generate the [Product] in high-fashion editorial style
with professional model and aspirational setting"

CATALOG:
"Show the [Product] on clean white background with soft
shadow, catalog photography style"

BILLBOARD:
"Create dramatic hero shot of [Product] with strong
visual impact and minimal distractions"

BROCHURE:
"Generate the [Product] in lifestyle context that tells
a story about its use and benefits"

Part 7: Flash vs Pro for Products

Feature Comparison

Feature	Flash (Standard)	Pro
Resolution	1K (1024px)	1K, 2K, 4K
Reference images	3 max	14 max
Text rendering	Basic	Advanced SOTA
Web grounding	No	Yes
Speed	Fast (~3-5 sec)	Quality-optimized (~8-15 sec)
Best for	Volume, prototypes	Final assets, print

When to Use Flash (Standard)

Ideal for:

✅ Rapid prototyping and ideation
✅ High-volume social media content
✅ Quick A/B testing variations
✅ Web-only imagery (1K sufficient)
✅ Budget-conscious campaigns
✅ Real-time/interactive applications
✅ Initial concept validation

Use Flash when:

You need many variations quickly
Images are for web/social (not print)
You're testing different scenes/concepts
Budget or speed is a primary concern

When to Use Pro

Ideal for:

✅ Print materials (catalogs, magazines)
✅ Large format displays (billboards)
✅ Images with text overlays
✅ Infographics with product
✅ Premium advertising campaigns
✅ Final production assets
✅ High-detail product showcase

Use Pro when:

Print resolution is required (2K+)
Text needs to be rendered on image
Maximum quality is essential
Creating final campaign assets
Detail visibility is critical

Resolution Considerations

Output Size	Minimum Resolution	Recommended Model
Social media (web)	1K	Flash
E-commerce listing	1K	Flash
Email marketing	1K	Flash
Presentation slides	1K-2K	Flash/Pro
Print catalog (small)	2K	Pro
Magazine ad	2K-4K	Pro (4K)
Billboard/large format	4K	Pro (4K)
Packaging design	4K	Pro (4K)

Reference Image Limits

Model	Max References	Impact
Flash	3 images	Good for simple products
Pro	14 images	Better for complex products

When more references matter:

Products with multiple angles needed
Complex products with many details
Products where color accuracy is critical
Items with unique textures/materials

Cost Analysis

Factor	Flash	Pro
Speed	~3-5 seconds	~8-15 seconds
API calls	Lower cost	Higher cost
Volume suitable	High volume OK	Use selectively
Quality	Good	Excellent

Strategy recommendation:

PROTOTYPE PHASE → Use Flash for iterations
TESTING PHASE → Use Flash for A/B variants
PRODUCTION PHASE → Use Pro for final assets

Part 8: Advanced Techniques

Multi-Product Scenes

Generate scenes with multiple products together:

PRODUCT BUNDLE:
"Generate the Ceramic Coffee Mug and Wooden Phone Stand
together in a gift box with tissue paper"

COMPLEMENTARY PRODUCTS:
"Show the Blue Gaming Mouse with matching keyboard and
mousepad in a gaming setup"

PRODUCT LINE:
"Generate all three color variants of the Water Bottle
arranged artistically on white background"

Tips for multi-product scenes:

Use landscape aspect ratio for multiple products
Ensure products don't overlap awkwardly
Create each product reference separately first
Describe spatial relationships clearly

Product + Character Combinations

Combine products with saved characters:

MASCOT WITH PRODUCT:
"Generate Banana Bob mascot character holding the
Premium Water Bottle, promotional style"

LIFESTYLE WITH CHARACTER:
"Show Explorer Emma character using the Wireless
Earbuds while hiking"

BRAND STORY:
"Generate Felix the Fox character showcasing the
Ceramic Coffee Mug in cozy cafe setting"

Style Transfer

Apply consistent visual styles across product images:

Create a brand style preset:

Create style preset "Brand Lifestyle":
- Prompt: "warm tones, natural lighting, premium lifestyle
  photography, magazine quality, aspirational setting"
- Negative: "cold colors, harsh shadows, amateur, cluttered"
- System: "Create images with consistent brand aesthetic"

Apply to all product generations:

Generate the [Product] in kitchen setting,
style_preset_name: "Brand Lifestyle"

Color Consistency

Maintain accurate product colors across scenes:

In scene prompt:

"Generate the [Product], maintaining exact original
colors as shown in reference image"

In additional details:

additional_details: "accurate color representation,
true-to-life colors, no color shifting"

In negative prompt:

negative_prompt: "wrong colors, color shift,
oversaturated, desaturated, color cast"

Lighting Matching

Match lighting across a series:

Consistent lighting setup:

Generate the [Product A] on white background with
soft studio lighting from upper left

Generate the [Product B] on white background with
soft studio lighting from upper left

Generate the [Product C] on white background with
soft studio lighting from upper left

Use style presets for consistency:

Create style preset "Studio Standard":
- Prompt: "professional studio setup, soft box lighting
  from upper left, clean white background"
- System: "Maintain consistent lighting across all images"

Background Continuity

Create a series with matching backgrounds:

CAMPAIGN SERIES:
"Generate [Product A] on rustic wooden table by sunny window"
"Generate [Product B] on the same rustic wooden table by sunny window"
"Generate [Product C] on the same rustic wooden table by sunny window"

Be specific about background details:

Scene: "on a weathered oak table with visible wood grain,
positioned by a large window with sheer white curtains,
warm afternoon sunlight streaming in, potted succulent
visible in background"

Conversational Editing for Products

After generating, refine with continue_editing:

Step 1: Generate the Premium Headphones on a desk
Step 2: Add a coffee cup in the background
Step 3: Make the lighting warmer
Step 4: Add subtle shadows under the product

Best practices:

Make one change at a time
Be specific about what to add/change
Use continue_editing for 80%+ correct images
Regenerate only if major changes needed

Part 9: Platform-Specific Guidelines

Instagram

Feed Posts (1:1 Square):

aspect_ratio: "1:1"
Scene: Product centered with breathing room
Style: Clean, bright, Instagram aesthetic
Tips: Leave space for text overlays if needed

Stories/Reels (9:16 Vertical):

aspect_ratio: "9:16"
Scene: Vertical composition, product in center/lower third
Style: Dynamic, lifestyle-focused
Tips: Consider safe zones for text/stickers

Example:

Generate the Premium Water Bottle in a fitness setting,
aspect_ratio: "1:1",
additional_details: "bright natural lighting, clean
Instagram aesthetic, minimalist composition"

Standard Pin (2:3):

aspect_ratio: "2:3"
Scene: Vertical lifestyle, product prominent
Style: Aspirational, beautiful, "pinnable"
Tips: Rich colors and beautiful settings perform well

Idea Pin (9:16):

aspect_ratio: "9:16"
Scene: Full vertical, storytelling composition
Style: Tutorial or inspirational format

Example:

Generate the Ceramic Coffee Mug in a cozy reading nook,
aspect_ratio: "2:3",
additional_details: "warm, inviting, Pinterest-worthy
aesthetic, soft lighting, lifestyle inspiration"

Facebook Ads

Feed Ad (1:1 or 4:5):

aspect_ratio: "1:1" or "4:5"
Scene: Product hero shot, clean background
Style: Professional, conversion-focused
Tips: Leave space for text overlay

Video Thumbnail (16:9):

aspect_ratio: "16:9"
Scene: Wide product shot
Style: Eye-catching, scroll-stopping

Example:

Generate the Blue Gaming Mouse hero shot,
aspect_ratio: "1:1",
additional_details: "professional e-commerce style,
clean background, product as focal point with space
for text overlay"

Amazon Product Images

Main Image Requirements:

aspect_ratio: "1:1"
Background: Pure white (#FFFFFF)
Product: Fills 85% of frame
Style: Clean, professional, accurate
No: Props, text, watermarks, accessories

Example:

Generate the [Product] on pure white background,
aspect_ratio: "1:1",
additional_details: "Amazon listing style, product fills
85% of frame, pure white background, no shadows,
professional product photography"

Lifestyle Images (Secondary):

Generate the [Product] in lifestyle setting showing use,
aspect_ratio: "1:1",
additional_details: "contextual lifestyle photo,
product in real-world use"

Shopify/E-commerce

Product Page Hero:

aspect_ratio: "1:1"
Style: Clean, professional, consistent
Background: White or brand-appropriate

Collection Page:

aspect_ratio: "1:1" (consistent across collection)
Style: Matching lighting and composition
Background: Consistent across all products

Example:

Generate the [Product] for Shopify listing,
aspect_ratio: "1:1",
additional_details: "e-commerce optimized, clean white
background, product centered, professional lighting,
web-ready quality"

Email Marketing

Header Image (16:9):

aspect_ratio: "16:9"
Scene: Wide hero shot
Style: Attention-grabbing, brand-consistent
Tips: Consider text placement areas

Product Feature (1:1 or 4:3):

aspect_ratio: "1:1"
Scene: Focused product shot
Style: Clean, click-worthy

Example:

Generate the [Product] newsletter header,
aspect_ratio: "16:9",
additional_details: "email marketing optimized,
eye-catching hero shot, space for headline overlay,
brand-consistent aesthetic"

Print Catalogs

Full Page (3:4 or 4:3):

aspect_ratio: "3:4" (portrait) or "4:3" (landscape)
Resolution: Use Pro model for 2K+
Style: High-quality, print-ready

Half Page:

aspect_ratio: "16:9" or "1:1"
Resolution: Pro model 2K minimum
Style: Editorial quality

Example:

Generate the [Product] for print catalog,
aspect_ratio: "3:4",
additional_details: "print-quality, high resolution,
editorial style, magazine-worthy composition,
professional product photography"

Part 10: Troubleshooting Guide

Product Not Found

Error:

Error: Product "Blue Mouse" not found

Causes:

Typo in product name
Product was deleted
Wrong tenant/API key
Product name is case-sensitive

Solutions:

Step 1: List all products to find exact name
"List all my saved products"

Step 2: Use exact name as shown
"Generate the Blue Gaming Mouse RGB on desk"

Step 3: Alternatively, use product ID
"Generate with product_id: 5 on desk"

Inconsistent Colors

Problem: Product appears with wrong colors in generated images

Causes:

Source image has color cast
Scene lighting conflicts
Background colors bleeding
AI interpretation variance

Solutions:

1. Check source image:
   - Ensure no color cast in original
   - Re-photograph with neutral lighting if needed

2. Add color accuracy to prompt:
   additional_details: "maintain exact product colors,
   accurate color representation, true-to-life colors"

3. Add to negative prompt:
   negative_prompt: "wrong colors, color shift,
   oversaturated, color cast"

4. Use neutral scene lighting:
   "Generate with soft neutral studio lighting"

Product Too Small/Large

Problem: Product is too small to see or dominates the frame

Too small:

Modify scene_prompt: "close-up of the product as main
focus, product prominent in frame, product fills 60%
of image"

Too large:

Modify scene_prompt: "product in context of larger
scene, environmental shot with product visible,
product occupies 30% of frame"

Background Overwhelming Product

Problem: Scene elements distract from the product

Solutions:

1. Add to negative prompt:
   negative_prompt: "cluttered background, distracting
   elements, competing focal points, busy scene"

2. Modify additional details:
   additional_details: "product as main focus, simple
   clean background, minimal distractions, product
   stands out"

3. Simplify scene:
   Change from: "on a kitchen counter with fruit,
   appliances, and dishes"
   To: "on a clean marble counter with soft lighting"

Wrong Lighting

Problem: Lighting doesn't match desired mood

Solutions:

1. Be specific about lighting:
   "Generate with soft diffused studio lighting
   from upper left"

2. Reference lighting style:
   "Professional product photography lighting setup"

3. Add to additional details:
   additional_details: "soft box lighting, even
   illumination, professional studio setup"

4. Negative prompt for unwanted lighting:
   negative_prompt: "harsh shadows, dark,
   underexposed, overexposed, flat lighting"

Blurry Product Details

Problem: Product appears blurry or lacks detail

Solutions:

1. Check source image quality:
   - Must be at least 1000x1000 pixels
   - Should be in focus, high resolution

2. Request sharp focus:
   additional_details: "sharp focus on product,
   high detail, crisp and clear"

3. For print quality, use Pro model:
   "Generate with Pro model at 2K resolution"

4. Negative prompt:
   negative_prompt: "blurry, soft focus, low detail,
   out of focus"

10-Point Debugging Checklist

When generation results aren't meeting expectations:

□ 1. SOURCE IMAGE: Is source at least 1000x1000?
□ 2. BACKGROUND: Is source on clean background?
□ 3. PRODUCT NAME: Is exact name used?
□ 4. SCENE PROMPT: Is scene clearly described?
□ 5. LIGHTING: Is lighting specified?
□ 6. ASPECT RATIO: Is correct ratio set?
□ 7. NEGATIVE PROMPT: Are issues addressed?
□ 8. ADDITIONAL DETAILS: Is style specified?
□ 9. MODEL: Is Flash/Pro appropriate?
□ 10. STYLE PRESET: Is brand consistency applied?

Part 11: Tools Reference

create_product_reference

Save a product image from a URL for marketing generation.

Parameters:

Parameter	Type	Required	Description
`product_name`	string	Yes	Unique name (1-100 chars)
`product_url`	string	Yes	Direct URL to product image
`product_description`	string	No	Product details for AI context
`tags`	string	No	Comma-separated tags

Example:

json

{
  "tool": "create_product_reference",
  "params": {
    "product_name": "Premium Water Bottle",
    "product_url": "https://example.com/bottle.jpg",
    "product_description": "32oz stainless steel with bamboo cap",
    "tags": "drinkware,outdoor,eco-friendly"
  }
}

Response:

Created product reference "Premium Water Bottle"

ID: 10
- Stored in R2
- Dimensions: 1200x1200
- Size: 245 KB
- Format: JPEG
- Tags: drinkware, outdoor, eco-friendly

Use with: generate_with_product

list_product_references

Browse saved product references.

Parameters:

Parameter	Type	Required	Description
`search`	string	No	Search name, description, tags
`page`	number	No	Page number (default: 1)
`page_size`	number	No	Results per page (default: 20)

Example:

json

{
  "tool": "list_product_references",
  "params": {
    "search": "bottle",
    "page_size": 10
  }
}

get_product_reference

Get full details of a specific product.

Parameters:

Parameter	Type	Required	Description
`id`	number	Either	Product ID
`product_name`	string	Either	Product name

Example:

json

{
  "tool": "get_product_reference",
  "params": {
    "product_name": "Premium Water Bottle"
  }
}

delete_product_reference

Remove a product reference from your library.

Parameters:

Parameter	Type	Required	Description
`id`	number	Either	Product ID to delete
`product_name`	string	Either	Product name to delete

Example:

json

{
  "tool": "delete_product_reference",
  "params": {
    "product_name": "Old Test Product"
  }
}

Note: Deleting removes the stored image. Generated marketing images are NOT deleted.

generate_with_product

Generate marketing images featuring a saved product.

Parameters:

Parameter	Type	Required	Description
`product_id`	number	Either	Product ID to use
`product_name`	string	Either	Product name to use
`scene_prompt`	string	Yes	Scene/setting description
`additional_details`	string	No	Extra styling details
`aspect_ratio`	string	No	Image aspect ratio
`negative_prompt`	string	No	What to avoid
`number_of_images`	number	No	Number to generate (1-4)
`style_preset_name`	string	No	Apply a saved style preset

Example:

json

{
  "tool": "generate_with_product",
  "params": {
    "product_name": "Premium Water Bottle",
    "scene_prompt": "on a rustic wooden table by sunny window",
    "additional_details": "warm atmosphere, lifestyle photography",
    "aspect_ratio": "landscape",
    "number_of_images": 2
  }
}

Internal prompt construction:

"Using the provided reference image showing [product_name],
generate [scene_prompt]. Maintain the exact design,
colors, materials, textures, and proportions of the
product. [additional_details]"

REST API Quick Reference

List Products:

bash

GET /api/products
curl -H "X-API-Key: sk_live_xxx" \
  "https://gobananasai.com/api/products?page=1&page_size=20"

Get Product:

bash

GET /api/products/:id
curl -H "X-API-Key: sk_live_xxx" \
  "https://gobananasai.com/api/products/3"

Create Product:

bash

POST /api/products
curl -X POST -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{"product_name":"...", "product_url":"..."}' \
  "https://gobananasai.com/api/products"

Update Product:

bash

PUT /api/products/:id
curl -X PUT -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{"product_description":"...", "tags":"..."}' \
  "https://gobananasai.com/api/products/3"

Delete Product:

bash

DELETE /api/products/:id
curl -X DELETE -H "X-API-Key: sk_live_xxx" \
  "https://gobananasai.com/api/products/3"

Generate with Product:

bash

POST /api/products/:id/generate
curl -X POST -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{"scene_prompt":"...", "aspect_ratio":"16:9"}' \
  "https://gobananasai.com/api/products/3/generate"

Batch Generate:

bash

POST /api/products/batch-generate
curl -X POST -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{"product_ids":[1,2,3], "scene_prompt":"..."}' \
  "https://gobananasai.com/api/products/batch-generate"

Product Statistics:

bash

GET /api/products/stats
curl -H "X-API-Key: sk_live_xxx" \
  "https://gobananasai.com/api/products/stats"

Part 12: Use Case Gallery

Use Case 1: E-commerce Product Catalog

Scenario: Create consistent product imagery for an online store.

Workflow:

STEP 1: Save all products
────────────────────────
Save product "Blue Widget" from https://...
Save product "Red Widget" from https://...
Save product "Green Widget" from https://...

STEP 2: Create consistent hero shots
──────────────────────────────────────
Generate the Blue Widget on white background, studio lighting
Generate the Red Widget on white background, studio lighting
Generate the Green Widget on white background, studio lighting

STEP 3: Create lifestyle shots
──────────────────────────────
Generate the Blue Widget on modern desk with laptop
Generate the Red Widget on modern desk with laptop
Generate the Green Widget on modern desk with laptop

STEP 4: Create size reference shots
────────────────────────────────────
Generate the Blue Widget next to credit card for scale

Scenario: Launch a new product with coordinated social content.

Workflow:

STEP 1: Save product
────────────────────
Save product "New Launch Product" from https://...

STEP 2: Create Instagram content
────────────────────────────────
Generate for Instagram feed (1:1):
"Generate New Launch Product in aesthetic flat lay with
accessories, aspect_ratio: 1:1"

Generate for Instagram Story (9:16):
"Generate New Launch Product being unboxed, vertical
composition, aspect_ratio: 9:16"

STEP 3: Create Pinterest content
────────────────────────────────
"Generate New Launch Product in aspirational lifestyle
setting, aspect_ratio: 2:3"

STEP 4: Create Facebook ad
──────────────────────────
"Generate New Launch Product hero shot with space for
text, aspect_ratio: 1:1"

Use Case 3: Seasonal Marketing

Scenario: Create seasonal variations for ongoing marketing.

Workflow:

STEP 1: Save evergreen product
──────────────────────────────
Save product "Core Product" from https://...

STEP 2: Generate seasonal variations
─────────────────────────────────────
SPRING:
"Generate Core Product surrounded by fresh flowers and
pastel colors, spring morning atmosphere"

SUMMER:
"Generate Core Product at beach scene with sand and
ocean waves in background"

FALL:
"Generate Core Product with autumn leaves and warm
orange tones, cozy feeling"

WINTER:
"Generate Core Product on festive holiday background
with pine branches and soft snow"

Use Case 4: A/B Testing Variations

Scenario: Create variations to test marketing effectiveness.

Workflow:

STEP 1: Save product
────────────────────
Save product "Test Product" from https://...

STEP 2: Generate variations
───────────────────────────
LIGHTING TEST:
A: "Generate with warm lighting"
B: "Generate with cool lighting"
C: "Generate with dramatic shadows"
D: "Generate with soft, even lighting"

BACKGROUND TEST:
A: "Generate on white background"
B: "Generate on lifestyle setting"
C: "Generate on gradient background"
D: "Generate floating with shadow"

ANGLE TEST:
A: "Generate straight-on product shot"
B: "Generate slight angle, 3/4 view"
C: "Generate top-down flat lay"
D: "Generate low angle, heroic perspective"

Use Case 5: Brand Style Consistency

Scenario: Maintain brand consistency across all product images.

Workflow:

STEP 1: Create brand style preset
─────────────────────────────────
Create style preset "Brand Standard":
- Prompt: "warm tones, natural lighting, premium feel"
- Negative: "harsh shadows, cold colors, amateur"
- System: "Consistent brand aesthetic"
- Aspect ratio: "1:1"

STEP 2: Save all products
─────────────────────────
Save product "Product A" from https://...
Save product "Product B" from https://...
Save product "Product C" from https://...

STEP 3: Generate with brand consistency
───────────────────────────────────────
Generate Product A in lifestyle setting,
style_preset_name: "Brand Standard"

Generate Product B in lifestyle setting,
style_preset_name: "Brand Standard"

Generate Product C in lifestyle setting,
style_preset_name: "Brand Standard"

Use Case 6: Multi-Channel Assets

Scenario: Create assets for multiple marketing channels.

Workflow:

STEP 1: Save product
────────────────────
Save product "Channel Product" from https://...

STEP 2: Generate for each channel
─────────────────────────────────
AMAZON (1:1, white bg):
"Generate on pure white background, 85% frame fill,
Amazon listing style, aspect_ratio: 1:1"

INSTAGRAM (1:1, lifestyle):
"Generate in lifestyle flat lay, Instagram aesthetic,
aspect_ratio: 1:1"

EMAIL HEADER (16:9):
"Generate hero shot with text space, email marketing
optimized, aspect_ratio: 16:9"

PINTEREST (2:3):
"Generate in aspirational setting, Pinterest-worthy,
aspect_ratio: 2:3"

FACEBOOK AD (4:5):
"Generate product focus with clean background,
ad-optimized, aspect_ratio: 4:5"

PRINT CATALOG (3:4):
"Generate editorial style, print-ready quality,
Pro model for 2K resolution, aspect_ratio: 3:4"

Appendix A: Parameter Reference

Scene Prompt Parameters

Parameter	Type	Required	Values
`scene_prompt`	string	Yes	Any descriptive text
`additional_details`	string	No	Styling, mood, quality
`negative_prompt`	string	No	What to avoid
`aspect_ratio`	string	No	See below
`number_of_images`	number	No	1-4
`style_preset_name`	string	No	Saved preset name
`style_preset_id`	number	No	Saved preset ID

Supported Aspect Ratios

Ratio	Orientation	Common Use
`1:1`	Square	Social media, e-commerce
`2:3`	Portrait	Pinterest pins
`3:2`	Landscape	Photography
`3:4`	Portrait	Catalogs
`4:3`	Landscape	Presentations
`4:5`	Portrait	Instagram
`5:4`	Landscape	Print
`9:16`	Tall portrait	Stories, Reels
`16:9`	Wide landscape	Video, banners
`21:9`	Ultra-wide	Banners

Product Reference Fields

Field	Type	Required	Description
`product_name`	string	Yes	Unique identifier (1-100)
`product_url`	string	Yes	Direct image URL
`product_description`	string	No	Product details
`tags`	string	No	Comma-separated

Image Formats Supported

Format	Extension	Notes
PNG	.png	Best for transparency
JPEG	.jpg, .jpeg	Smaller file size
WebP	.webp	Modern, efficient

Appendix B: Prompt Templates Library

E-commerce Templates

HERO SHOT:
"Generate [Product] on pure white background with
professional studio lighting, soft shadows, product
centered and filling 80% of frame"

LIFESTYLE:
"Show [Product] in natural lifestyle setting with
complementary props, warm lighting, magazine quality"

DETAIL:
"Close-up of [Product] showing texture and material
quality, dramatic side lighting, high detail"

SCALE:
"Generate [Product] next to [common object] for size
reference on white background"

INSTAGRAM FEED:
"Generate [Product] in aesthetic flat lay arrangement,
bright natural lighting, Instagram-worthy composition,
aspect_ratio: 1:1"

INSTAGRAM STORY:
"Show [Product] in vertical lifestyle shot, dynamic
composition, story-optimized, aspect_ratio: 9:16"

PINTEREST:
"Generate [Product] in aspirational lifestyle setting,
rich colors, beautiful environment, aspect_ratio: 2:3"

FACEBOOK AD:
"Create [Product] hero shot with clean background,
space for text overlay, professional quality,
aspect_ratio: 1:1"

Campaign Templates

SEASONAL - SPRING:
"Generate [Product] surrounded by fresh flowers,
pastel colors, spring morning light, renewal theme"

SEASONAL - SUMMER:
"Show [Product] in bright summer setting, beach or
outdoor scene, vibrant and energetic atmosphere"

SEASONAL - FALL:
"Generate [Product] with autumn leaves, warm orange
and brown tones, cozy harvest feeling"

SEASONAL - WINTER:
"Create [Product] in festive holiday setting, pine
branches, snow accents, warm and inviting"

Professional Photography Templates

STUDIO WHITE:
"Generate [Product] on pure white (#FFFFFF) background,
soft box lighting from upper left, minimal shadow,
clean professional studio photography"

STUDIO GRADIENT:
"Show [Product] on soft gray gradient background,
professional lighting setup, product photography style"

EDITORIAL:
"Generate [Product] in high-end editorial setting,
dramatic lighting, premium magazine quality,
aspirational lifestyle"

MINIMALIST:
"Create [Product] floating against clean white
background with subtle shadow, minimalist aesthetic,
plenty of negative space"

Appendix C: Troubleshooting Decision Tree

ISSUE: Generation results not meeting expectations

├── Is product appearing correctly?
│   ├── NO: Product distorted or wrong
│   │   ├── Check source image quality (>1000px?)
│   │   ├── Check source background (clean?)
│   │   └── Try simpler scene prompt
│   │
│   └── YES: Product OK, but...
│       │
│       ├── COLORS wrong?
│       │   ├── Add "maintain exact product colors"
│       │   ├── Add negative "wrong colors, color shift"
│       │   └── Check source image color accuracy
│       │
│       ├── PRODUCT too small?
│       │   └── Add "product as main focus, prominent"
│       │
│       ├── PRODUCT too large?
│       │   └── Add "product in context of scene"
│       │
│       ├── BACKGROUND distracting?
│       │   ├── Simplify scene prompt
│       │   ├── Add negative "cluttered, distracting"
│       │   └── Add "clean background, minimal"
│       │
│       ├── LIGHTING wrong?
│       │   ├── Specify lighting in prompt
│       │   └── Add negative for unwanted lighting
│       │
│       ├── STYLE inconsistent?
│       │   ├── Use style preset
│       │   └── Add detailed style instructions
│       │
│       └── QUALITY low?
│           ├── Use Pro model for higher resolution
│           ├── Add "high detail, professional quality"
│           └── Add negative "blurry, low quality"

STILL HAVING ISSUES?
├── List products to verify exact name
├── Check product reference details
├── Try regenerating with simplified prompt
├── Use 10-point debugging checklist
└── Contact support with image examples

Appendix D: Glossary

Term	Definition
Product Reference	A saved product image in R2 storage with metadata in D1 database
R2	Cloudflare's object storage (similar to S3) where product images are stored
D1	Cloudflare's SQL database where product metadata is stored
MCP	Model Context Protocol - the interface for AI tool interaction
Scene Prompt	The description of the setting/environment for the product
Additional Details	Extra styling instructions added to generation
Negative Prompt	Instructions on what to avoid in generation
Aspect Ratio	The width:height proportion of the generated image
Style Preset	A saved combination of prompt, negative, and style settings
Tenant	An account/organization with isolated data and settings
Flash (Standard)	The faster, more economical Gemini model
Pro	The higher-quality Gemini model with more features
Reference Image	The source product image used as visual reference
Fidelity	Accuracy of product representation in generated images
Hero Shot	Primary product image, typically on clean background
Lifestyle Shot	Product shown in real-world context/use
Flat Lay	Top-down product arrangement photography
Batch Generation	Creating multiple images for multiple products at once

Quick Reference Card

┌─────────────────────────────────────────────────────────────────┐
│                    PRODUCT MARKETING QUICK REFERENCE            │
├─────────────────────────────────────────────────────────────────┤
│                                                                  │
│  SAVE PRODUCT:                                                   │
│  Save product "[Name]" from [URL]                               │
│  with description "[desc]" and tags "[tags]"                    │
│                                                                  │
│  GENERATE:                                                       │
│  Generate the [Product Name] [scene description]                │
│  aspect_ratio: "[ratio]"                                        │
│  additional_details: "[styling]"                                │
│  negative_prompt: "[avoid]"                                     │
│                                                                  │
│  COMMON ASPECT RATIOS:                                          │
│  • 1:1   - Square (Instagram, Amazon)                          │
│  • 9:16  - Vertical (Stories, Reels)                           │
│  • 16:9  - Wide (Banners, Video)                               │
│  • 2:3   - Portrait (Pinterest)                                │
│                                                                  │
│  QUALITY TIPS:                                                   │
│  • Source image: 1000px+ resolution                            │
│  • Clean background on source                                   │
│  • Specific scene descriptions                                  │
│  • Use style presets for consistency                           │
│  • Pro model for print quality                                  │
│                                                                  │
│  TOOLS:                                                          │
│  • create_product_reference                                     │
│  • list_product_references                                      │
│  • get_product_reference                                        │
│  • delete_product_reference                                     │
│  • generate_with_product                                        │
│                                                                  │
│  REST API: /api/products                                        │
│                                                                  │
└─────────────────────────────────────────────────────────────────┘

Last updated: 2024 | Go Bananas! Product Marketing Master Guide v1.0

Product Marketing Master Guide ​

Table of Contents ​

Part 1: Foundations ​

What is the Product Reference System? ​

Why Product Marketing Matters ​

The Problem Without Product References ​

How It Works (Architecture Overview) ​

Key Concepts: Products vs Characters ​

Part 2: Quick Start ​

3-Step Workflow ​

Your First Product ​

Your First Marketing Image ​

Behind the Scenes ​

Part 3: Source Image Quality ​

Ideal Product Images ​

Resolution Requirements ​

Background Best Practices ​

Lighting Guidelines ​

Product Positioning ​

Common Image Problems ​

Source Image Checklist ​

Image Format Comparison ​

Part 4: Creating Product References ​

From URL (Step-by-Step) ​

Method 1: MCP Tools (Conversational) ​

Method 2: REST API ​

Product Naming Conventions ​

Writing Effective Descriptions ​

Tagging Strategy ​

Multiple Product Variants ​

Updating Products ​

Part 5: Generating Marketing Images ​

Basic Generation ​

Scene Prompts That Work ​

Additional Details ​

Aspect Ratios by Platform ​

Negative Prompts ​

Style Preset Integration ​

Batch Generation ​

Part 6: Product Photography Scenarios ​

E-commerce Hero Shots ​

Lifestyle Photography ​

Detail/Close-up Shots ​

Size Comparison Images ​

Environment Variations ​

Seasonal Campaigns ​

Social Media Content ​

Print Advertising ​

Part 7: Flash vs Pro for Products ​

Feature Comparison ​

When to Use Flash (Standard) ​

When to Use Pro ​

Resolution Considerations ​

Reference Image Limits ​

Cost Analysis ​

Part 8: Advanced Techniques ​

Multi-Product Scenes ​

Product + Character Combinations ​

Style Transfer ​

Color Consistency ​

Lighting Matching ​

Background Continuity ​

Conversational Editing for Products ​

Part 9: Platform-Specific Guidelines ​

Instagram ​

Pinterest ​

Facebook Ads ​

Amazon Product Images ​

Shopify/E-commerce ​

Email Marketing ​

Print Catalogs ​

Part 10: Troubleshooting Guide ​

Product Not Found ​

Inconsistent Colors ​

Product Too Small/Large ​

Background Overwhelming Product ​

Wrong Lighting ​

Blurry Product Details ​

10-Point Debugging Checklist ​

Part 11: Tools Reference ​

Product Marketing Master Guide

Table of Contents

Part 1: Foundations

What is the Product Reference System?

Why Product Marketing Matters

The Problem Without Product References

How It Works (Architecture Overview)

Key Concepts: Products vs Characters

Part 2: Quick Start

3-Step Workflow

Your First Product

Your First Marketing Image

Behind the Scenes

Part 3: Source Image Quality

Ideal Product Images

Resolution Requirements

Background Best Practices

Lighting Guidelines

Product Positioning

Common Image Problems

Source Image Checklist

Image Format Comparison

Part 4: Creating Product References

From URL (Step-by-Step)

Method 1: MCP Tools (Conversational)

Method 2: REST API

Product Naming Conventions

Writing Effective Descriptions

Tagging Strategy

Multiple Product Variants

Updating Products

Part 5: Generating Marketing Images

Basic Generation

Scene Prompts That Work

Additional Details

Aspect Ratios by Platform

Negative Prompts

Style Preset Integration

Batch Generation

Part 6: Product Photography Scenarios

E-commerce Hero Shots

Lifestyle Photography

Detail/Close-up Shots

Size Comparison Images

Environment Variations

Seasonal Campaigns

Social Media Content

Print Advertising

Part 7: Flash vs Pro for Products

Feature Comparison

When to Use Flash (Standard)

When to Use Pro

Resolution Considerations

Reference Image Limits

Cost Analysis

Part 8: Advanced Techniques

Multi-Product Scenes

Product + Character Combinations

Style Transfer

Color Consistency

Lighting Matching

Background Continuity

Conversational Editing for Products

Part 9: Platform-Specific Guidelines

Instagram

Pinterest

Facebook Ads

Amazon Product Images

Shopify/E-commerce

Email Marketing

Print Catalogs

Part 10: Troubleshooting Guide

Product Not Found

Inconsistent Colors

Product Too Small/Large

Background Overwhelming Product

Wrong Lighting

Blurry Product Details

10-Point Debugging Checklist

Part 11: Tools Reference