OpenAI's new image tool upends Amazon product photography economics
OpenAI's new AI image-generation tool, launched last week, is rapidly changing product photography for Amazon sellers, removing the need for pricey agencies and putting high-quality, targeted visuals within everyone's reach. One seller's standard operating procedure for combining AI data analysis and image generation to build image sets tailored to a single ideal customer persona went viral, drawing more than 11,000 comments.
Early testers who generated more than 500 images with the feature have distilled a set of practical lessons. Prompt like a creative director, specifying the scene, mood, layout, backdrop, and lighting rather than just naming the product. Skip plain white backgrounds in favor of lifestyle backdrops and soft, engaging surfaces. Pair features with benefits, showing why a feature matters. Use modular layouts that combine hero shots with callouts and icons.
Further tips: specify soft natural or studio-style lighting; be specific rather than merely realistic, asking for crisp shadows, matte textures, or a product held in hand; place logos in the upper third with clearly defined font styles and a strong call to action; favor visual hierarchy over minimalism, centering the product and framing it with key benefits; and tell a story with relatable people, clear usage scenarios, or small cues like motion or half-opened packaging.
Separately, research shows most Amazon sellers remain single-country operators: 69 percent sell exclusively on one marketplace, only 6 percent have expanded to six or more, and fewer than 1 percent are genuinely global across 11 or more countries. One reason is market size: the 22 non-US Amazon marketplaces together barely exceed in traffic that of Amazon's US site alone.
Amazon is also restructuring fees across Lightning Deals, Best Deals, Prime Exclusive Discounts, and Coupons taking effect June 2, 2025, a move drawing complaints across the seller community. Sellers of products under $22 stand to save significantly on coupons and gain flexibility for shorter promotions, while sellers of higher-priced or high-volume products face steeper costs. One example shows a high-performing Best Deal rising from a $300 flat fee to $990 under the new daily-fee-plus-one-percent-of-sales formula, an increase of more than 230 percent.
Critics note Amazon repeatedly frames fee hikes as performance-based improvements, using incremental adjustments like variable percentage fees to quietly raise costs, especially for high-volume sellers. Recommendations include ramping coupon usage for low-priced items, bundling to raise average order value, running promotions more selectively on high-priced products, adjusting 2025 promotional budgets, and remembering that stacking deals with coupons incurs cumulative fees.
Deal data adds context: the average Lightning Deal discount runs 19.6 percent and Prime Exclusive Deals 16.9 percent, Chinese sellers dominate deal volume at 77.6 percent of Lightning Deals and 79.4 percent of Prime Exclusive Deals, and some brands successfully run always-on promotions for visibility.
In tools news, a new all-in-one creative canvas called Flora is gaining attention for stitching text, image, and video AI models into a single visual workspace, letting sellers prototype product images, video ads, and brand messaging without coding or switching between separate tools. Sellers are also sharing techniques for enriching back-end listing attributes, a data approach said to perform well with Amazon's Rufus AI assistant.
In other headlines: comparison images are being credited with lifting listing performance, sellers lose an estimated $1.2 billion a year to recoverable marketplace mistakes, Amazon is extending Prime Day 2025 to four days, and TikTok's potential US ban remains a political flashpoint. Organizers of the Iceland seller conference say attendees represent more than $3.5 billion in annual sales.
