How to Generate On-Brand Product Photos Without a Designer: 5 Nano Banana 2 Pro Workflows Small Businesses Stole from Figma's New AI Stack

You need product photos for your Instagram. Your designer is on vacation. Your phone camera makes everything look like a garage sale listing. And Canva's stock photos scream "generic small business energy."

Sound familiar?

Figma just quietly rolled out stronger AI image tools across their entire suite (May 29–31, 2026), and small businesses immediately started using them to generate on-brand visuals without bouncing between twelve different apps. The secret? Reference images + detailed prompts + the right AI model.

Here's the thing: You don't need Figma to steal their workflow. Nano Banana 2 Pro on soracai.com gives you the same reference-image power with better control over aspect ratios and output quality. Let me show you exactly how five small businesses are using this workflow right now.

What You Actually Need (No Designer Required)

Before we dive in, grab these:

Pro tip: Figma's update lets non-designers work directly in design files. We're doing them one better—you'll generate ready-to-post images in the exact aspect ratio your platform needs (9:16 for TikTok, 4:5 for Instagram feed, 16:9 for your website hero).

Workflow #1: The Coffee Shop Menu Board

The Problem: You change seasonal drinks weekly. Hiring a photographer each time costs $200-400.

Step 1: Gather Your Reference Images

Head to soracai.com/create and upload 2-3 photos:

Nano Banana 2 Pro accepts up to 5 reference images to guide generation. More references = more consistent brand look.

Step 2: Write a Detailed Prompt

Bad prompt: "Pumpkin spice latte professional photo"

Good prompt: "Top-down view of a ceramic mug filled with pumpkin spice latte, cinnamon stick garnish, scattered coffee beans on rustic wooden table, warm morning sunlight from left, shallow depth of field, Fujifilm color palette, cozy autumn cafe aesthetic"

Why it works: You're giving the AI lighting direction, composition, mood, and technical specs. This is exactly what Figma users discovered—detail wins.

Step 3: Choose Your Aspect Ratio

For Instagram Stories: 9:16
For Instagram feed: 4:5
For your website: 16:9

Nano Banana 2 Pro offers 11 aspect ratios. Pick before generating—cropping later kills quality.

Step 4: Enable PRO Mode

Click Nano Banana 2 PRO mode. Yes, it's 4 coins instead of 1, but the difference between standard and PRO is the difference between "nice try" and "wait, you hired a photographer?"

PRO mode gives you:

Step 5: Generate and Tweak

Hit generate. Wait 2-3 minutes. If the first result is 80% there, regenerate with a tweaked prompt rather than settling. Small businesses using this workflow report generating 3-4 options and picking the best takes 15 minutes total.

Real result: A Portland coffee shop now creates all their seasonal drink photos this way. Cost per image: $0.40 (4 coins). Previous photographer cost: $250 per shoot.

Workflow #2: The Boutique Product Lineup

The Problem: You sell handmade jewelry. You need lifestyle shots for each piece, but models + photographers = $$$$.

The Hack:

Upload a photo of your actual product + 2-3 reference images of the lifestyle aesthetic you want ("minimalist hand jewelry on marble surface with morning light").

Prompt example: "Close-up of silver crescent moon necklace draped over a woman's collarbone, soft window light from right, cream silk fabric background, editorial jewelry photography style, Vogue aesthetic, sharp focus on pendant"

Choose 4:5 for Instagram feed or 1:1 for a grid layout.

Appspace (the digital signage platform) just added end-to-end AI image generation in their editor for exactly this use case—small retailers spinning up localized promo images without a designer. You're doing the same thing, but for social media instead of in-store screens.

Pro tip: For jewelry and small products, add "macro photography" or "product photography 85mm lens" to your prompt. It tells the AI to nail that shallow depth-of-field look.

Workflow #3: The Gym's Social Media Blitz

Reface app just added AI hairstyles (May 27, 2026) for virtual makeovers. Gyms and fitness studios are catching on—but instead of hairstyles, they're generating motivational training content.

The Setup:

Upload reference images of your gym's interior + equipment. Prompt for action shots:

"Athletic woman performing kettlebell swing in modern gym, motion blur on kettlebell, dramatic side lighting, concrete walls, motivational fitness photography, Nike ad aesthetic, powerful and dynamic"

Aspect ratio: 9:16 for TikTok and Reels.

Generate 5-7 variations with different exercises (deadlift, battle ropes, box jumps). Now you have a week of content.

Bonus move: Once you have static images, head to soracai.com/ai-dance and turn one into a dancing video. Yes, a kettlebell doing the robot is ridiculous. Yes, it'll get 10x more engagement than another "motivation Monday" post. Kling 2.6 motion control makes it smooth enough that people will watch the whole thing.

Cost: 8 coins for the dance video. Your TikTok going viral: priceless.

Workflow #4: The Restaurant's Daily Special

Appspace's new AI image feature (mid-May 2026) lets restaurants generate "2-for-1 lunch special today" graphics for in-store screens. You're doing the same for Instagram Stories.

The Speed Run:

Real talk: A taco truck in Austin does this every morning. Their "today's special" Stories get 3x more replies than their regular posts because the photos look professionally shot.

Workflow #5: The "Fake It Till You Make It" Product Launch

You're launching a new product line. You have prototypes but not final units. You need marketing assets now.

The Pre-Launch Play:

Upload photos of your prototype + reference images of the aesthetic you're targeting.

Prompt: "Professional product photography of [your product], studio lighting, white seamless background, commercial advertising style, high-end retail aesthetic, sharp focus, Canon 5D Mark IV look"

Generate in multiple aspect ratios:

Nuwacom just added FLUX models and Imagen 4 for business workspaces (late May 2026), turning generic platforms into content studios. You're doing the same thing, but soracai.com's 11 aspect ratio options mean you're not cropping and losing quality.

Pro tip: Generate your hero image in 21:9 (cinematic ultrawide) for website banners. It's the ratio nobody thinks to use, and it looks expensive.

Pro Tips That Separate Amateurs from "Wait, You Hired Who?"

Tip #1: Reference Images Are Your Cheat Code

Figma's update made reference images the star feature. Use all 5 slots:

Tip #2: Steal Prompts from the Library

Head to soracai.com/prompts—there are 1000+ curated prompts organized by category. Find one close to your need, copy it, customize the details. Why start from scratch?

Tip #3: Batch Your Generations

Don't generate one image at a time. Queue up 5-10 variations in one session. Tweak prompts slightly (change lighting, angle, mood). Pick the best 2-3. You'll find your "formula" faster.

Tip #4: PRO Mode for Anything Customer-Facing

Standard mode (1 coin) is fine for internal mockups. Nano Banana 2 PRO mode (4 coins) is non-negotiable for:

The color accuracy and detail difference is immediately obvious. Don't cheap out on the stuff customers see.

Tip #5: Combine with Video for Maximum Reach

Once you have your hero image, turn it into a video:

Static images get scrolled past. Moving content stops thumbs.

Troubleshooting: When Your AI Images Look Like AI Images

Problem: Colors look oversaturated and fake.
Fix: Add "natural color grading" or "Fujifilm color palette" to your prompt. Specify "subtle" and "realistic" lighting.

Problem: Composition feels off or cluttered.
Fix: Use photography terms: "shallow depth of field," "rule of thirds," "negative space," "centered composition." The AI understands camera language.

Problem: Product details are wrong (wrong color, weird proportions).
Fix: Upload a clear reference image of your actual product. In your prompt, emphasize "product accuracy" and "commercial photography."

Problem: Every generation looks different (inconsistent branding).
Fix: Use the same 3-5 reference images for all generations. Create a "prompt template" with your brand descriptors (color palette, lighting style, mood) and only change the product/subject.

Problem: It's taking forever to generate.
Fix: Peak usage times (lunch, evenings) can slow things down. Generate in batches during off-hours, or queue multiple at once.

The Bottom Line: Designer-Quality Photos in 15 Minutes

Figma's AI update made waves because it solved the "context switching" problem—designers no longer need to jump between tools. Small businesses have the same problem, but worse: you're the designer, marketer, and photographer.

Nano Banana 2 Pro on soracai.com gives you:

The five workflows above are working right now for coffee shops, boutiques, gyms, restaurants, and product launches. Total time per image: 10-15 minutes. Total cost: $0.10-$0.40 depending on mode.

Your designer is still on vacation. But your Instagram is about to look like you hired an agency.

Try it: soracai.com/create

P.S. Once you've got your product photos sorted, circle back and turn one into a dancing video at soracai.com/ai-dance. Trust me on this. A dancing coffee cup gets more engagement than 47 perfectly composed static shots. The algorithm is weird like that.