Back to Resources

What Amazon requires of a main image, how to build the rest of the stack around the objections that stop a sale, where video fits, what Amazon’s AI disclosure rule asks of you, and how visual search changes what a good product photo is.

A converting Amazon image stack in 2026 starts with a main image that passes Amazon’s rules and still wins the click on a phone: a pure white background, the product filling 85% to 100% of the frame, and no text, logos, or props. Behind it, Amazon staff recommend “at least six images and one video,” and each of those secondary images should answer one specific reason a shopper might not buy, in the order shoppers tend to raise them.

AI-generated images don’t hurt you on their own. Amazon has no ban on them, and it builds AI image generation into its own ad and A+ Content tools. What Amazon asks for is accuracy and, since July 2026, disclosure of photorealistic AI-generated people. AI images cause trouble when the product in the picture doesn’t match the one in the box, and that shows up in returns and reviews.

Most image stacks we review were planned as a photo shoot. The ones that convert were planned as a sales conversation, one objection per frame.

The Main Image Has Two Jobs

The main image has to clear Amazon’s review and then beat every other thumbnail on the search results page. Those are different standards, and most brands stop at the first one.

What Amazon requires. Per Amazon staff guidance on the Seller Forums:

• A “pure white (RGB 255, 255, 255)” background
• The product filling “85-100% of the image frame”
• No “text, graphics, logos, watermarks, or promotional messaging on main images”
• No props or items that don’t come with the product
• “500 to 10,000 pixels” on the longest side, with at least 1,000 pixels for zoom
• JPEG, TIFF, PNG, or non-animated GIF

What wins the click. Compliance gets you listed. The click comes from how the thumbnail reads at phone size, next to the listings around it. A few things move it:

• Fill the frame. Push toward 100%. A product floating in white space looks smaller than the competitor next to it.
• Pick the angle that explains the product. A three-quarter view usually shows shape and depth better than a flat front shot. For packs, show the full count so the shopper sees the value.
• Match the search. If shoppers search for the product by a feature (a lid, a handle, a set of three), make that feature visible in the thumbnail.
• Check it where shoppers see it. Search your main keyword on your phone and look at your listing in the grid. If you can’t tell what it is in two seconds, neither can the shopper.

The main image is also the easiest place to lose sales you never see. A low click-through rate never generates a complaint or a bad review, so it can sit unnoticed for months while the traffic goes to the listing next to yours.

Build the Secondary Images Around Objections

Every secondary image should answer one question that stops a sale. Shoppers have already told you what those questions are: in your one- to three-star reviews, your competitors’ low reviews, the questions on your detail page, and your return reasons. Rank the objections by how often they come up, then assign one image to each.

A sequence that works for most products:

1. The core benefit. What the product does for the shopper, shown rather than claimed.
2. Size and scale. The product next to a hand, a common object, or a person, with dimensions. In our experience, size is behind more “not as expected” returns than any other detail.
3. Feature proof. A close-up of the detail that justifies the price: the stitching, the seal, the motor, the material.
4. In use. A lifestyle image of the product in a realistic setting, used by the person who’s likely to buy it.
5. The top objection. Compatibility, care, assembly, safety, or whatever your reviews say most often, answered visually.
6. What’s in the box. Everything the shopper receives, laid out. This closes the gap between expectation and delivery.

Text is allowed on secondary images, and it helps when it’s short. Treat each image like a billboard: one idea, a few words, type big enough to read on a phone. If a claim on an image needs a footnote, it probably needs a rewrite. Our guide to optimizing Amazon listings for mobile covers what changes on the small screen.

If other sellers offer your product, you also need to know whose images Amazon shows. Since 2024, Amazon has said it “will prioritize brand owner images” when several sellers contribute images to the same listing, starting with hardlines and expanding from there. If a reseller’s images are showing on your listing, our guide to dealing with Amazon hijackers covers how to report detail page changes.

Where Video Fits

Video is open to more sellers than many brands assume. Amazon says “both brand registered and non-brand registered sellers with at least three months of selling history” can add shoppable videos, and you can add up to four per ASIN. Amazon’s internal data from January to June 2026 shows that “product detail pages with their first shoppable video added saw an average 24% increase in sales.” The category figures run higher: 32.7% in Toys, 30.5% in Apparel, 26.7% in Home, and 25.8% in Beauty.

Amazon says “most successful videos are just 30 to 90 seconds long.” In our experience, the videos that earn that lift show the thing a still image can’t: how the product moves, folds, pours, attaches, or feels in use. A slideshow of your existing photos set to music adds little.

Amazon’s own guidance holds a tradeoff. Amazon staff recommend at least six images plus a video. Amazon’s video page says that “for shoppable videos to appear in the main media block of product detail pages, the product must have fewer than six images.” With six or more images, the video won’t appear in the gallery where shoppers swipe first.

Our take: if your product needs to be seen in motion to make sense, keep the gallery at five images so the video stays in the main block. If your product is easy to understand from stills, use all your image slots and treat video as a supporting asset. When you can’t tell which group you’re in, test it.

Do AI-Generated Images Hurt?

Not by themselves. Amazon has no general ban on AI-generated images, and it ships its own generators: an image generator for ads since October 2023, and AI image tools inside A+ Content Manager since November 2025. The posts claiming Amazon “banned AI images” don’t trace back to an Amazon source.

Two rules do apply:

• Accuracy. Amazon’s guidance is to “ensure images accurately represent your product.” An AI image that changes the product’s color, proportions, texture, or included parts breaks that rule, whatever tool made it.
• Disclosure for AI people. In July 2026, Amazon told sellers that images showing photorealistic AI-generated people need to be disclosed with a “contains-synthetic-performer” tag in the image’s metadata. The rule covers listing images and A+ Content.

In practice, AI is useful for the scene around the product and risky for the product itself. A generated backdrop around a real photo of your product is low risk and fast. A generated product is where errors creep in: a handle in the wrong place, a finish that looks more premium than it is, a strap that doesn’t exist. Those errors cost you twice, once in returns and again in reviews that say “doesn’t look like the picture.”

Our rule of thumb: photograph the real product, use AI for context, and check every final image against a physical unit before it goes live. The same applies to A+ modules, which our guide to Amazon A+ Content covers in detail.

How Visual Search Changes the Brief

Shoppers can now start an Amazon search with a camera instead of a keyword. Amazon launched Lens Live in September 2025, describing it as “a deep learning visual embedding model to match the customer’s view against billions of Amazon products.” In June 2026, Amazon added Shop by Style, Visual Suggestions, and Circle to Search to Lens. Alexa for Shopping, which launched in May 2026, also supports “snapping a photo to use visual search.”

Amazon hasn’t said how its visual matching uses your listing images, so treat any claim about that with suspicion. What it has said is that a shopper’s photo gets matched against products. That’s a reasonable case for what good images already do: show the product clearly, from more than one angle, in its true color and shape, without filters or stylized renders that make it look like something else.

The practical change is in the brief to your photographer. Ask for at least one clean shot from each angle a shopper might photograph the product in the real world, the way it looks on a shelf, in a hand, or in use. Our Alexa for Shopping guide covers how the assistant changes the rest of the listing.

Test Before You Reshoot

Opinions about images are cheap. Manage Your Experiments lets you test them on real traffic. It covers product images, titles, bullet points, descriptions, and A+ Content, and it requires a Brand Representative role in Brand Registry, a Professional selling plan, and enough traffic on the ASIN to produce a result. Amazon’s guidance puts a typical test at 8 to 10 weeks.

Start with the main image, since it controls the click and every other image depends on the shopper getting to the page. Change one thing per test (angle, fill, pack count shown) so you know what moved the result. Once the main image is settled, test the image that answers your top objection. Our Amazon listing optimization guide covers how images fit with titles, bullets, and keywords in the rest of the listing.

How Canopy Builds Image Stacks

Our brand managers and creative team build image stacks from the same data that drives the rest of the account: reviews, return reasons, search terms, and conversion by ASIN. We map objections first and write a shot list with one job per image before anyone picks up a camera. Our product photography and video team handles the shoot, and because the same brand manager runs your advertising, they see whether the new images move click-through and conversion, then test the next change in Manage Your Experiments.

Canopy is a full-service omnichannel marketing agency. Our partners have generated $3.3 billion in revenue, with an average 84% year-over-year profit increase and 99.1% partner retention.

Frequently Asked Questions