An E-commerce Product-Photography Playbook with Floniks

A mid-range product photography studio session — half a day, one photographer, basic equipment rental — typically costs several hundred to over a thousand dollars before post-processing. For a catalog of 50 SKUs, each requiring on-white hero shots, lifestyle scenes, and at least one variant colorway, the bill runs into the tens of thousands. That price point is reasonable for a funded brand doing a seasonal catalog shoot, but it is prohibitive for a solo Amazon seller with 200 products, a Shopify brand launching weekly drops, or a marketplace reseller who needs images for products arriving daily.

Floniks' /ai-image and /product-design tools, combined with the multi-step workflow editor, replace the studio for the majority of product photography use cases. The key insight is that most e-commerce imagery follows one of four predictable templates: (1) on-white isolation shots, (2) contextual lifestyle scenes, (3) colorway/variant renders, and (4) ad-ready composites. Each template maps to a repeatable AI workflow that you build once and run indefinitely across your catalog.

This playbook is structured around those four templates, with concrete prompt patterns, workflow setups, and quality-control checkpoints at each stage. By the end, you will have a system that produces marketplace-compliant imagery — meeting Amazon’s white-background and minimum-resolution requirements, for example — at a fraction of traditional studio cost, and in hours rather than days.

Marketplace platforms — Amazon, eBay, Etsy, Walmart Marketplace — typically require a pure white background (RGB 255,255,255) for the primary listing image, with the product filling at least 85 % of the frame. Getting this right in AI generation requires explicit prompt engineering, not just adding "white background."

Base prompt skeleton for on-white shots:

"Product photography, [product name and material description], isolated on pure white background, soft box studio lighting from 45° camera left with fill light camera right, no cast shadows, no reflections on background, sharp focus throughout, commercial product photo, 3:4 aspect ratio"

The key phrase is "no cast shadows, no reflections on background" — without it, AI tends to generate attractive shadow gradients that look great in lifestyle contexts but fail marketplace compliance checks. Add "fill light camera right" to eliminate the hard-edged shadow that a single-source setup creates.

For product categories with complex surfaces — glassware, metallics, patent leather — add material-specific lighting notes: "broad back-light to reveal glass transparency" for glassware, "specular highlight along top edge" for metallic finishes, "matte lighting, no hot spots" for matte surfaces.

After generation, run a quick compliance check: zoom to 200 % in any image viewer and inspect the four corners. Pure white (#FFFFFF) in all four corners confirms the background meets marketplace standards. If you see a slight gray cast, note the prompt adjustment (increase "pure white" emphasis or add "overexposed background") for the next run.

Lifestyle photography — your candle on a marble bathroom counter, your backpack against a mountain trail, your skincare product beside a sunny window — traditionally requires location scouting, prop sourcing, and model fees on top of studio costs. AI generation collapses this into a prompt and a product reference image.

The most reliable workflow for lifestyle product placement uses Floniks' /editor to run a two-node pipeline: (1) an image-conditioning node that takes your existing on-white product shot as the reference and (2) a scene generation node that places the conditioned product into a described environment. This approach preserves the product’s actual appearance — the exact colors, logo placement, and silhouette from your real product photo — while generating an entirely new background environment.

Lifestyle prompt anatomy:

"[Product name] placed on [surface material and color], [scene description], [ambient light source], [background depth and blur], [color temperature], [scene mood], commercial lifestyle photography"

Example: "Matte black travel mug placed on weathered pine picnic table, autumn forest in background, dappled morning sunlight filtering through trees, shallow depth of field, warm amber tone, outdoor lifestyle photography, 4:5 aspect ratio"

Aim to generate three to five scene variants per product — indoor and outdoor, aspirational and practical, seasonal. Buyers make purchasing decisions based on imagining the product in their own life, so the more contexts you cover, the wider the audience you speak to. Use the batch workflow to run all five scene variants from one session rather than generating each individually.

If your product comes in six colorways, you do not want to photograph six separate samples. Traditional photography requires physical samples for every variant — a logistical and financial burden that delays catalog launches and makes it expensive to test new colors before committing to inventory.

Floniks' batch variant workflow solves this with a color-transfer pipeline in /editor: (1) generate (or upload) your best-looking hero shot of the base color, (2) pass it through a color conditioning node with a palette descriptor for each variant, (3) generate all variants in parallel. The pipeline preserves the product’s shape, texture, and lighting while changing only the color.

Color descriptor examples for precise variant generation:

• "deep navy blue, matte finish, same texture and form as reference"
• "forest green, semi-gloss, consistent highlight placement"
• "ivory white, exact same silhouette and shadow placement as reference"

For textured products — knitwear, leather goods, wood grain items — also specify texture preservation: "same knit texture and stitch pattern, color changed to heather gray". Without this, the AI may simplify the texture when applying the new color.

After generating variants, do a side-by-side consistency review: line up all colorways at equal size and confirm the product silhouette, shadow length, and background match across all variants. Inconsistencies in shadow or background treatment signal that a variant was generated from a different effective prompt interpretation and will look mismatched in a grid display on your product listing page.

Product imagery for ads must satisfy multiple technical specifications simultaneously: Facebook Feed requires 1:1 or 4:5, Stories require 9:16, Google Shopping requires 1:1 with a 10 % safe zone margin, Amazon Sponsored Products have their own aspect ratio guidelines. Manually cropping or resizing a single hero shot for each placement loses critical product detail and often looks amateurish — but re-generating per placement is expensive if done manually.

The solution is a multi-format ad pipeline in Floniks' /editor with parallel output branches. Start with your approved lifestyle or on-white hero shot as the anchor. Route it into three or four branches, each with a format-specific composition instruction:

• "Square 1:1, product centered, clear negative space above and below for text overlay"
• "Vertical 4:5, product in lower two-thirds, open space in top third for headline text"
• "Story 9:16, full-bleed lifestyle scene, product positioned center-screen, large enough to read at phone size"

Add a text safe zone reminder in your composition prompts: products that bleed into the top 15 % of a Story frame get covered by the username overlay; products in the bottom 20 % get covered by the call-to-action button. Composing deliberately around these zones in the generation prompt is far more reliable than trying to reposition in post.

For seasonal campaign bursts — Black Friday, Valentine’s Day, back-to-school — create a campaign variant workflow that injects a seasonal background cue (falling leaves, snow, confetti) into your standard lifestyle template. This extends an evergreen product image into a timely ad creative with minimal additional generation cost.

AI-generated product imagery at scale requires a structured review process to catch issues before images go live. A marketplace listing with an obviously AI-artifact background (warped text, morphed reflections, floating objects) damages brand credibility and can trigger platform content reviews. A four-gate review process applied to every image before publishing catches the vast majority of quality issues.

Gate 1 — Silhouette integrity: Does the product’s outline match the actual product? Look for missing handles, merged layers between product and background, or distorted packaging text. Packaging text and logos are the most common AI failure mode on product images; if the logo is important to the listing, mask it in post or use a real product photo as the conditioning reference with the logo already in position.

Gate 2 — Background compliance: Zoom to 200 % on all four corners. No gray cast, no gradient, no stray objects. For lifestyle images, confirm the background is plausible — no impossible lighting angles, no physically absurd scene elements.

Gate 3 — Variant consistency: In a grid view, do all colorway variants look like they were shot in the same session? Shadow length, background tone, and product angle should be identical across variants. Inconsistency here stands out immediately in the listing’s image carousel.

Gate 4 — Platform resolution: Export at a minimum of 2000 px on the long edge (Amazon recommends 2560 px). AI-generated images at their native output resolution usually meet this threshold; downscaled or heavily cropped images may not. Check file size too — images under 1 MB at listing dimensions sometimes indicate quality compression that will show artifacts on high-DPI displays.

The ultimate goal of this playbook is not to make individual product images — it is to build a system that can onboard new SKUs and generate a complete image set in under 30 minutes per product. Here is the catalog workflow structure that makes this possible.

Step 1 — Create a master product template in /editor. This workflow accepts three inputs: (a) a product category tag (e.g., "home decor", "apparel", "electronics"), (b) a product description with material, color, and key features, and (c) an optional reference product photo. Output nodes generate: hero white background, two lifestyle scenes, and a square ad creative.

Step 2 — Build a batch input sheet. Each row represents one SKU: product name, category, description, variant colors. New products are added as rows; the batch workflow processes all new rows in a single run.

Step 3 — Run weekly or on-demand. For daily drops (common in reselling and wholesale categories), run the batch at end of day before publishing. For a planned catalog refresh, run the full batch on a production schedule aligned with your listing update cycle.

Step 4 — Archive approved outputs. Store approved images with a filename convention that includes SKU, image type, and format (e.g., sku-1234_hero_1x1.jpg). This makes it trivial to pull any image for an ad campaign, email template, or social post without re-generating.

A well-structured catalog workflow in Floniks functions as a scalable visual production line — the kind of infrastructure that previously required a dedicated studio team, and now runs on a laptop with an afternoon of workflow setup.