Visual Mastery with Nano Banana: How We Generate Studio-Quality Product Shots Without a Studio

Let's talk about the thing that kills most small Indian product businesses before they even launch.

Not the product. Not the pricing. Not the supply chain.

The photos.

You've seen it happen. A genuinely good product — well-made, well-priced, actually useful — sitting on a grimy table under a tubelight, photographed in portrait mode on someone's two-year-old phone. The listing gets zero clicks. Not because the product is bad. Because it looks bad, and in e-commerce, looking bad is the same thing as being bad.

We almost made this exact mistake with Stashed.

Why Product Photography Is the #1 Blocker for Small Indian Businesses

Here's the economics. A half-decent product photographer in any Indian city charges somewhere between ₹3,000 and ₹15,000 per product shoot. A fashion/lifestyle photographer who can do atmospheric, brand-level shots? You're looking at ₹20,000–₹50,000 for a day, not counting studio rental, props, and models.

For a brand that hasn't sold a single unit yet, that's an absurd ask.

So what happens? One of three things:

1. You spend the money, blow your early budget, and pray it works.
2. You take bad photos yourself and launch looking like a side project.
3. You don't launch at all because it "doesn't look ready."

Option 3 is more common than anyone admits. We know because we almost did it.

We had Stashed ready — the design, the Velcro panel concept, the first sample bags — but we hadn't started manufacturing yet. We had nothing to photograph. And we needed to test demand, build a waitlist, and start talking about the brand. The full story of building Stashed covers how we went from zero to a functioning brand — this post focuses on the visual side of that journey.

That's when we went fully into Nano Banana Pro.

What Is Nano Banana and Why It Works Better Than Midjourney for Product Shots

Nano Banana Pro is an AI image generation tool. If you've heard of Midjourney, it's in the same category — but the reason we use it instead of Midjourney for product work comes down to one thing: control over product fidelity.

Midjourney is incredible for artistic, atmospheric images. It's brilliant at vibes. But when you need a specific bag, in a specific colorway, with specific stitching detail, on a specific surface — Midjourney tends to hallucinate its own version. It interprets your prompt rather than executing it.

Nano Banana Pro handles structured product prompts better. When we describe a matte black crossbody bag with a flat Velcro front panel and minimalist design, it renders that bag — not an artistic interpretation of "bag energy."

For brands working with real products, that distinction matters enormously.

The Anatomy of a Great Product Photography Prompt

This is the part most people skip. They write "bag on white background" and wonder why it looks like a stock photo from 2009.

A great product photography prompt has five components:

1. The Subject — what the product is, its key features, its material and finish.

2. The Lighting — this is the biggest lever. Soft diffused light looks professional. Harsh direct light looks amateur. Golden hour light looks aspirational. State it explicitly.

3. The Angle — front, three-quarter, overhead (flatlay), close-up detail. Pick one. Don't let the model decide.

4. The Background — white seamless, textured concrete, warm wood, lifestyle scene, outdoor environment. Again: be specific.

5. The Mood — this is the only place where "feeling" language is allowed, and even here, anchor it to concrete references. "Editorial, like a Vogue India feature" is better than "luxurious."

Most people write prompts with only the subject. That's why most results are average.

5 Real Prompt Examples That Work for Indian Products

These are prompts we've tested. Copy and refine them.

For a bag (like Stashed):

"Studio product photography, matte black rectangular crossbody bag with flat Velcro front panel, minimalist design, no visible branding, photographed at a three-quarter angle, soft diffused light from the left, warm off-white textured background, sharp focus on stitching and zipper detail, editorial quality, high resolution"

For a kurta:

"Flatlay product photography, handwoven cotton kurta in deep indigo, neatly folded with collar visible, natural morning light from above, placed on raw linen fabric background, wooden block print texture visible on fabric edge, warm tones, clean composition, no wrinkles, fashion editorial style"

For jewellery:

"Macro product photography, 22K gold temple jewellery necklace, placed on dark polished marble surface, single soft spotlight from upper right, reflections visible in marble, extreme close-up showing craftsmanship detail, no background clutter, jewellery catalogue style, crisp shadows"

For food:

"Restaurant food photography, stainless steel plate with sambar rice and three side dishes, photographed from 45-degree angle, natural daylight from window on left, banana leaf underneath the plate, rustic wooden table surface, warm South Indian kitchen atmosphere, steam effect, close crop, editorial quality"

For electronics:

"Clean product photography, wireless earbuds in open charging case, placed on matte grey textured surface, overhead flat lay, perfectly symmetrical composition, soft even studio lighting, no shadows, tech product catalogue style, white case with chrome accents in sharp focus"

Before and After: What a Bad Prompt Produces vs. a Refined One

Here's an actual comparison we ran.

Bad prompt:

"A bag on a nice background"

Result: Midjourney gave us a brown leather tote on a stock-photo marble surface. Looked like a generic Amazon listing. Zero brand identity.

Refined prompt:

"Product photography, matte black urban crossbody bag, flat Velcro front panel, three-quarter angle view, soft natural light from upper left, warm concrete wall background with subtle texture, shallow depth of field on front panel detail, editorial fashion aesthetic, no text or branding"

Result: Something we actually used in our early Instagram posts. The angle looked intentional. The lighting felt real. The texture on the bag was visible.

The difference isn't artistic talent. It's specificity.

How We Generated Stashed's Launch Photos Before Manufacturing Started

This is the part that still kind of amazes us.

We had two physical samples. That's it. They were sitting in our first-floor room above the jewellery shop. No studio. No model. No photographer.

We photographed the samples on a plain white bedsheet in our room with the window open for natural light. Decent reference photos — functional, but not launch-worthy.

Then we ran those reference images through Nano Banana Pro with lifestyle scene prompts. We described the setting we wanted: metro station background, young person carrying the bag, Mumbai-style crowded commute environment. We generated 40 variations in about two hours.

Out of 40, around 12 were genuinely usable. Three of them were strong enough that we used them as our pre-launch Instagram content. People in our DMs were asking where to buy the bag before we'd manufactured a single unit beyond those two samples.

That's demand validation. That's also AI compressing a ₹40,000 photography budget into an afternoon.

Tips: Specificity Beats Creativity Every Time

We want to say this directly because it runs counter to how most people think about creative tools.

When you're doing product photography, don't be creative with your prompt. Be surgical.

"Luxurious and modern" tells the model nothing. "Soft diffused studio lighting, two-light setup, warm key light from the right, cool fill from the left" tells it exactly what to render.

Every vague word in your prompt is a decision you're handing to the model. Sometimes that's fine. For product work, it almost never is.

Write your prompt like you're briefing a photographer who has never met you, doesn't know your brand, and will photograph exactly what you describe — nothing more, nothing less.

Common Mistakes and How to Fix Them

Mistake 1: Asking for multiple angles in one prompt. Fix: One prompt, one angle. Want three angles? Make three prompts.

Mistake 2: Over-describing the product's emotional benefit. "This bag represents freedom and movement" — the model doesn't care. Describe the object.

Mistake 3: Forgetting the surface. Background gets attention but the surface the product sits on is often more important for realism. Specify it.

Mistake 4: Not iterating. Your first generation won't be the best one. Run five variations. Pick the best. Refine from there. The refinement loop is where the quality comes from.

Mistake 5: Using it for products you haven't designed clearly. If you don't know exactly what your product looks like, Nano Banana will make one up. Get your design nailed first.

How to Batch Generate 20 Variations in 30 Minutes

This is the workflow we actually use:

1. Write your master prompt — the one that describes the product correctly.
2. Identify three variables you want to test: background, angle, lighting.
3. Create a 2×2×2 matrix. That's 8 prompt variations with one variable changed each time.
4. Run all 8. Look at the outputs.
5. Pick the top 2. Make 6 micro-variations of each (tweak one word at a time).
6. You now have 12 polished outputs from ~20 generations.
7. Total time: 25–35 minutes depending on generation speed.

This is not magic. It's a system. And systems are faster than inspiration.

The Bigger Picture

Product photography used to be a gatekeeping mechanism. If you didn't have the budget for a professional shoot, your product looked unprofessional, full stop.

That gate is gone now.

What replaced it is a different skill: the ability to write precise, structured prompts that communicate visual intent. That's learnable. It took us about three weeks of daily practice to get consistently good results.

The brands that figure this out first have a real competitive advantage — not just in cost savings but in the ability to iterate. We can test five different visual directions for Stashed in a day. A brand paying a photographer can test one per week. Nano Banana is just one part of our full AI stack — the other tools we run daily are covered in that post.

That speed compounds.

---

The full product photography workflow — including prompt templates for 15 product categories, our generation-to-selection system, and how to adapt outputs for Shopify, Instagram, and print — is in Phase 1 of 2BFT Academy.

Free to start. No pitch at the end. Just the actual workflow.

Start at 2bft.in/skills →