Claude prompts for Shopify product photos — the 8 shots you actually need

Claude prompts for Shopify product photos — 8 shots

Product photography is where DTC brands most consistently under-invest — then blame conversion. A great product page is 60% imagery, 30% copy, 10% social proof. Founders spend hours on descriptions and ship a Shopify page with 4 identical-angle white-background shots.

Claude can direct the shot planning and execute background swaps + lifestyle composites — but not by itself. Below is the workflow: the 8 shots every Shopify product page needs, the exact prompts we use to brief a photographer or generate AI shots, and where Claude replaces vs. augments human photography.

Table of Contents

Key Takeaways

Question Answer
What's the biggest imagery mistake? 4 identical-angle white-background shots + no context.
How many shots does a product page need? 8 minimum. 12 for premium categories.
Can Claude generate all of them? No. Some (scale, lifestyle-with-humans) still need real photography.
What CAN Claude do well? Background swaps, environmental composites, styled product-only shots, shot-list planning.
What tool actually generates the images? Claude briefs the shots; image models (Nano Banana Pro, Midjourney, Ideogram) render.
How much does this replace a photo shoot? Cuts a shoot by 40-60%. Doesn't eliminate it for hero + human shots.

Why 4 white-background shots aren't enough

Look at your product page. Now look at the last thing you bought online. The difference is usually the imagery.

The 4 white-background shots most Shopify stores ship answer one question: "what does the product look like." They fail to answer the questions that actually drive purchase:

Add these and conversion lifts 10-25% in most categories. The imagery IS the selling.

The 8 shots every product page needs

1. Hero (white or clean background, straight-on). The identifiable shot. This is the ad-set image, the OG image, the "what does the product look like" shot.

2. Scale shot. Product held in hand, or next to a reference object (mug, book, phone). Answers "how big is this."

3. Detail / construction close-up. Stitching, material texture, hardware, buttons. Answers "how is this made."

4. What's included. Product + accessories + packaging + care instructions laid out. Answers "what am I paying for."

5. Lifestyle / in-use. Product being used in its intended context. Answers "does this fit my life."

6. Environmental (product in setting). Product on a shelf, on a desk, in a room — no human. Answers "where does this belong."

7. Alternate angle. Back, side, or bottom of product. Answers "what does the whole thing look like."

8. Comparison / usage. With similar products (size chart), with previous version (if evolution matters), or a "before/after" if applicable.

Premium categories add:

9-10. Behind-the-scenes. Production shot, materials sourcing, artisan at work.

11. Video (5-15 seconds, silent, autoplay-friendly).

12. UGC-style shot (customer's real environment, phone-photo aesthetic).

Claude prompt — generate a full shot list

Feed Claude your product spec + brand and it generates a complete shoot brief:

BRAND: [name]
PRODUCT: [name]
CATEGORY: [category]
KEY MATERIALS: [list]
KEY FEATURES: [3-5 features]
TARGET CUSTOMER: [1-line]
BRAND VISUAL DIRECTION: [minimal / warm / editorial / lifestyle / studio / other]

I need a full shot list for this product's Shopify page.
Follow the 8-shot framework: hero, scale, detail, contents,
lifestyle, environmental, alt-angle, comparison.

For each of the 8 shots, give me:
1. Shot name
2. What the shot must clearly show (1-2 sentences)
3. Composition notes (angle, framing, lighting direction)
4. Props or context objects (be specific — brand, if it
   matters, or category)
5. Model/hand notes if applicable
6. Common mistake to avoid

Then rank the 8 shots by which are (a) required for launch and
(b) can wait for round 2.

Output: a shot list you can hand to a photographer OR feed into image-generation tools shot by shot.

Claude prompt — background swap on existing shots

You have decent product-only shots but ugly backgrounds. Claude + a modern image model (Nano Banana Pro, GPT-Image, etc.) can generate replacements without regenerating the product:

INPUT: [attach existing product shot]
PRODUCT: [name] — [1-line description]
DESIRED BACKGROUND: [describe scene — "warm morning kitchen, marble counter, natural side light from window on left, soft focus on cabinets in background"]
KEEP UNCHANGED: [product itself, its exact position, its lighting, its shadow direction]
STYLE: [photorealistic / editorial / lifestyle / minimalist]

Generate a new version of this image with the described background,
keeping the product exactly as it appears. The product must not be
re-rendered — only the background changes. Preserve the existing
lighting angle on the product.

Avoid: warping the product, changing the product's shape,
inconsistent shadows between product and new background.

Works for ~70% of product types. Fails on transparent products, chrome, and anything with complex reflections.

Claude prompt — lifestyle composite from product-only shots

Take a white-background shot and place it in a full scene:

INPUT: [attach white-background product shot]
PRODUCT: [name and use case]
SCENE: [describe the intended lifestyle context in detail — "a young professional's home office desk at 8am, coffee mug to the left, notebook to the right, natural light from a window behind the camera, warm neutral color palette"]
CAMERA: [angle — "eye-level 35mm-like, slight depth of field, product in sharp focus"]
STYLE: [aesthetic — "warm editorial, DTC brand photography style, Kinfolk-inspired"]
DO NOT: [invent a human, invent branded items, change the product]

Generate the image. If the product size in the composition looks
wrong, prioritize keeping the product's scale realistic.

Best for lifestyle-context shots where you don't need a human hand or face. Adding humans is where AI generation still frequently fails (see below).

Claude prompt — direct a real photographer

For hero, scale-in-hand, and human-in-use shots, real photography still wins. Claude can produce the brief:

BRAND: [name]
BRAND AESTHETIC: [3-5 reference images or a paragraph]
PRODUCT: [name + hero benefit]
SHOOT PURPOSE: [Shopify PDP + Meta ads + Instagram]

Generate a photographer shoot brief for a half-day shoot,
including:

1. Prop list (be specific — mug brand, notebook brand, plant
   type, fabric type for background)
2. Model brief (age range, aesthetic, wardrobe direction, how
   product should be held/worn)
3. Lighting setup (natural, artificial, direction, warmth)
4. 12-shot list with 3 hero shots, 4 detail shots, 3 lifestyle
   shots, 2 alt-angles
5. Post-production notes (color grade, retouching level, aspect
   ratio outputs needed)
6. What NOT to do (common shoot mistakes for this category)

Format as a clean brief the photographer can print and follow.

Output: a professional shoot brief. Save the founder from writing it themselves; save the shoot from missing shots.

Where Claude falls short

Be honest about the current failure modes of AI image generation:

1. Faces + hands. Still uneven. AI-generated hands touching product often look slightly wrong. Real photos win for hero + lifestyle-with-humans shots.

2. Text on packaging. AI models still garble product text 50%+ of the time. Real shots for any product where label legibility matters.

3. Reflective surfaces. Chrome, glass, glossy plastic — AI often gets reflections wrong (inconsistent with the environment).

4. Scale accuracy. Image models don't know your product is 6 inches vs. 6 feet. If scale is off, customers spot it immediately.

5. Brand-specific props. If your brand needs a specific reference object (a competitor's product for comparison, a very specific brand of coffee mug), AI can't reliably render it.

Real photography for hero + human-in-frame + label-visible shots. AI for background swaps + environmental composites + shot planning + alt-angles.

Talk to Branva

Product page imagery is the highest-leverage single fix on most low-converting Shopify stores. If your product page is 4 white-background shots, book a working session and we'll audit your imagery + show a mock-up of what the 8-shot framework looks like for one of your hero SKUs.

Frequently Asked Questions

What image model does Claude use for these?

Claude's own image capabilities are limited; the workflow above is Claude for planning + prompting, with the actual rendering handled by dedicated image models (Nano Banana Pro, Midjourney, Ideogram, GPT-Image, etc.). Claude generates the prompt; you paste into the image tool.

Can I skip real photography entirely?

For accessories, tech, home goods, food, and stationery — often yes. For apparel, beauty, jewelry, or anything requiring a person — no. Human-with-product shots still need real photography.

What resolution should I export at?

Shopify recommends 2048×2048 for product images (they compress and resize automatically). Export at that, use image_url: width: X in your theme to serve the right size per breakpoint.

Do AI-generated product photos hurt SEO?

No direct penalty from Google. Google Image Search may weight studio-shot originals higher when both are indexed. Practically: use AI for background swaps + environmental composites, keep hero and detail shots real.

How much can this replace an agency?

For a brand doing $50K+/mo, this workflow replaces roughly $2K-5K/mo of contracted product photography, keeping only the 1-2 hero shoots per quarter for human-in-frame work.

Related reading