A standard Amazon listing gives you seven image slots. Most of the listings we get handed to fix are using three. That gap is the cheapest conversion improvement available to a new seller, and almost nobody takes it. We manage stores doing $200k a month for paying clients, and when a launch underperforms, the gallery is the first place we look, before the bids and before the bullets. One client’s sports launch put four bundled items in the main image and held a 20% price premium at 25% ACoS in month one. Another client sold a generic product with clean, expensive photos and lost 12 months to it. Photos decide a lot. They don’t decide everything.
What Amazon actually gives you: 7 slots, or 6 plus a video
Your slot count depends on one thing, and it isn’t your budget. It’s whether you’re Brand Registered.
If you’re a standard seller without Brand Registry, you get seven image slots. The first one is your main image, the on-white shot that appears in search results. The other six are yours to use however you want: lifestyle shots, infographics, size comparisons, packaging, anything that sells the product.
If you’re Brand Registered, you get six image slots plus one video slot. The video replaces what would have been your seventh image. That’s the trade, and for most products it’s worth taking, though not always. I’ll come back to that later in the post.
Either way, the rule we score every client gallery against is the same one:
- A shopper should understand the product completely from the images alone, without reading a single word of your title, bullets, or description.
- Every slot you’re entitled to is filled. An empty slot is a free sales asset you left on the table.
- Each image does a different job. Seven angles of the same product on white is not a gallery, it’s one image repeated.
- The images survive being shrunk to a thumbnail on a phone screen. If the text in your infographic needs pinching to read, it isn’t doing its job.
Amazon’s own technical requirements sit underneath all of this. Images need to be at least 1,000 pixels on the longest side for the zoom function to switch on, and Amazon recommends 1,600 pixels or larger. Zoom matters more than it sounds, because it’s what lets a buyer inspect stitching, material, and finish before they commit. Check the live image guidelines in Seller Central before you brief a photographer, since Amazon adjusts these.
Why the gallery decides the sale before your copy gets a turn
Shopping on Amazon is a visual comparison, not a reading exercise. A buyer searches a keyword, sees a grid of near-identical thumbnails, and picks two or three to open. Your title helps you get into that grid. Your main image decides whether you get opened.
Here’s the clearest example from our own client work. A sports and fitness client bundled his core product with a hand band, a wrist band, a resistance band, and a small carrying bag. Instead of mentioning the bundle in the title and bullets like most sellers do, we put all four items in the main image. The buyer’s scan read “4 items” against a competitor’s “1 item” before they read a word. Month one came in at 25% ACoS and 13% TACoS, at a price 20% above the category average, with the listing holding first page. Click-through and conversion both ran above category norms. The premium price never scared anyone off, because the thumbnail had already made the value case.
Now the other side of it, because photography is not magic. Another client came to us two years ago with an existing wood therapy massage tool. Generic product, five or more competitors selling the same physical item. We ran ASIN-targeted ads at competitors with weak photography, refreshed his images, and reoptimized the listing on a weekly cycle in month one. Every tactic lifted sales for two to four weeks, then decay. After 12 months of lift and decline, we told him to pivot out of the product. We were winning the visual battle and losing the wallet battle, because when two products are interchangeable, buyers pick the cheaper one.
The lesson we carry from both launches: photography sells differentiation. It cannot invent it. If your product is identical to nine competitors, better photos buy you a slower decline, not a win. Sort the product out first, and our guide on how to differentiate a saturated niche covers the moves that actually change the outcome.
The 7 shots, in order
This is the sequence we brief for every managed client launch. Shoot them in this order too, because each one answers the question the previous image raised.
Shot 1: the main image
Pure white background at RGB 255, 255, 255. The real, physical product filling roughly 85% of the frame. No added text, badges, borders, or graphics. Amazon enforces this one, and a violation can get the listing suppressed.
Inside those constraints, the job is to make your differentiation visible. If you bundle, every bundled item goes in this shot. If your product comes in a premium gift box and competitors ship in poly bags, the box belongs here. If your version has a feature the others lack, frame it so the eye lands on it.
The accuracy rule is not optional. Whatever is in this image has to be what ships. Show a bundle the buyer doesn’t receive and you’ve bought yourself “item not as described” returns, a falling star rating, and eventually a suppressed listing.
Shot 2: what’s in the box
A flat lay of everything the buyer receives, laid out and separated so each piece is countable. Product, accessories, manual, storage case, spare parts.
This shot does two jobs. It confirms the bundle promise your main image made, and it pre-empts the most common pre-purchase question on Amazon, which is some version of “does it come with X”. Sellers who bundle and skip this shot are paying for the extra items in COGS, freight, and packaging complexity without getting the conversion credit for them.
Shot 3: scale and dimensions
A size reference, either against a human hand or body, or against a familiar object, with the actual measurements printed on the image.
Returns in a lot of categories are driven by the product being bigger or smaller than the buyer pictured. The dimensions are already in your bullets and your detail page, and buyers still don’t read them. Put the numbers on a picture instead. We treat this as a returns-prevention image as much as a conversion image, and returns are a P&L line just like ad spend.
Shot 4: the feature infographic
Your product in the frame with two to four features called out in short labels. Not eight. Two to four.
The temptation is to list every feature you have. Resist it, because the buyer is looking at this on a phone and small type at thumbnail size is just grey noise. Pick the features that separate you from the listing directly above you in search results. Big type, short phrases, one idea per callout.
Shot 5: the lifestyle shot
The product in the place it will actually live, used by the kind of person who will actually buy it.
This is the image that answers “is this for me”. A kitchen tool on a real counter in a real kitchen, not floating on a gradient. A fitness product mid-use, not sitting on a gym floor. If you have more than one buyer type, and most products do, the lifestyle shot is where you show the one with the most search volume behind it.
A detail that matters more than it should: match the setting to the buyer you priced for. If you sell at a premium, a scuffed budget kitchen in the background undercuts the price before anyone reads it. If you sell to a beginner, an expert setup makes the product look like it needs skills they do not have. The room is part of the pitch.
Shot 6: the objection killer
Every product has one thing that stops the sale. Find yours and photograph the answer.
For a food container it’s the leak test. For anything with a battery it’s the charge time or the port. For a cutting board it’s whether it warps in the dishwasher. Read your competitors’ one-star and three-star reviews and you’ll find the objection written out for you, repeatedly, by their own customers. Then build the image that removes it. This is the single most underused slot in the seven, and it’s usually the cheapest to shoot.
Shot 7: packaging, or the video
Standard sellers finish here with the packaging shot, especially if the product could be bought as a gift. Premium packaging justifies a premium price, and it only does that if the buyer can see it before they order.
This is one of the four differentiation moves we teach, and it is the one most often paid for and then hidden. If competitors ship in poly bags and you spent money on a gift-ready box, that spend earns you nothing until a buyer sees it. Which is why the packaging usually belongs in the main image as well, not only in slot seven. Gift buyers are a separate pool of demand inside most categories, and they self-select on exactly this image.
Brand Registered sellers trade this slot for a video. More on that below.
How we brief a gallery, and what it costs
The photos are the easy part. The brief is where launches are won or lost, because a photographer who doesn’t sell on Amazon has no idea which of your features matter.
Our process for a managed client runs like this:
- Pull the competitor grid. Screenshot the first page of search results for the main keyword. This is the visual field your thumbnail has to beat, and it tells you instantly what everyone else’s main image looks like. We paste that screenshot into ChatGPT or Gemini and ask what we could add or show that the visible competitors are not showing. The model is good at reading patterns across a grid of offers. It is bad at judging which of its ideas actually move buyers, so we filter every suggestion by hand and commit to one or two.
- Read competitor reviews for objections. The three-star reviews on the top four listings give you shot 6 for free.
- Find the competitor winning on fewer reviews. If a listing with 355 reviews is outselling one with more than 1,300, that seller is doing something right without the review advantage. Study their gallery specifically: what they lead with, what sits in slot two, what story the images tell. Then do the same thing better.
- Write one line per slot. Seven lines. Each line says what’s in the frame and what the shopper should understand after looking at it for one second. If you can’t write that line, the shot has no job and shouldn’t be taken.
- Shoot the samples you already ordered. We spend about $150 per supplier on samples and ship them to Pakistan rather than the US, because Pakistan is closer to China, samples land in about a week, and per-sample shipping costs less. Those samples are your photography subjects. You do not need production inventory to build a gallery.
- Build the gallery before inventory lands. The listing should be live and complete on the day your stock checks in, not two weeks after.
On cost, the honest numbers from our own launches. On a vegetable chopper launch we ran, photography plus listing content came to about $800. That covered a professional photographer plus a separate content creator who built the listing images, the A+ Content, and the brand story. That was the pre-AI-image era for us. Today the same workflow, AI-assisted and briefed properly, lands closer to $200 to $400 at similar quality. For context, that $800 sat inside an all-in launch of about $8,450 for that product, so photography was under 10% of the cash deployed.
On cadence, we optimize images weekly for the first month of a launch. Four cycles in 30 days, reviewing performance and changing image order, callout wording, and which shot sits in position two. After that, quarterly. Once a listing is ranking and converting, weekly tinkering introduces noise and risks killing whatever is working.
AI images: the three-way split that keeps you out of trouble
We use AI in our own client work, and this is the part sellers get wrong most often in 2026. Amazon does not have one rule about AI images. It has three different situations, and they have different answers.
Brand Registry and GTIN Exemption proof photos: real photos only. When you apply for Brand Registry or a GTIN Exemption, Amazon asks for photos of real packaging or a real product with your brand name physically printed on it, held in hand, shot from each side. Rendered mockups get rejected. Our workflow for every client is to ask the supplier, during the manufacturing window, to print the brand on actual packaging and photograph it from each angle. The supplier emails the photos, we submit them. That route has been accepted on every client filing we’ve run, and it avoids the one to two wasted application cycles that AI mockups cost.
Gallery and lifestyle images: AI is fine, if it’s accurate. This is the lower-risk use and it’s where most of the old photography cost disappeared. Generating your product in different settings, at different angles, with clean lighting, is acceptable as long as the result honestly represents the physical item. We use Google Labs for this.
The main image: AI only as cleanup, never as invention. Amazon’s policy for the main image is unchanged whether a human or a model made it. Real product, pure white background, roughly 85% of the frame, no graphics or text, accurate to what ships. Using AI to clean up or composite a real product shot is fine. Generating a main image of a product that doesn’t match your physical item is not, and it lands you in suppression and “item not as described” territory.
Keep those three straight and AI saves you several hundred dollars per launch with no downside. Blur them together and you’ll be appealing a suppressed listing during your launch window.
Video: when trading your 7th image is worth it
If you’re Brand Registered, the seventh slot can hold a video instead of an image. Amazon’s own published figure is that adding a video can raise conversion rate by up to 10%, and “up to” is doing work in that sentence, so treat it as an upper bound rather than a forecast.
Uploading it is straightforward. In Seller Central, hover the Catalog tab, click Upload and Manage Videos, then Upload a video. You supply four things: the video file, a clear title, every ASIN you want the video to appear on, and a thumbnail. The thumbnail is the part sellers rush. It’s the frame a shopper sees before they decide whether to press play, so it should show the product clearly and give a reason to watch. Amazon reviews the video against its guidelines before it goes live.
Our rule on whether to take the trade: video wins when the product needs a demonstration, and loses when it doesn’t. A folding stroller, a multi-step kitchen gadget, an assembly-required organizer, anything where “how does it actually work” is the blocker, should have the video. A straightforward product with an obvious use gets more from a seventh still image, usually the packaging shot or a second objection killer. Don’t upload a video because the slot exists.
One more thing worth knowing. Video sits in the image carousel, so it competes with your stills for attention rather than adding to them. That’s the real cost of the trade, and it’s why we decide it per product instead of applying a blanket rule.
The mistakes we flag most
Across 47 client onboardings since 2022, the same gallery problems keep showing up.
The invisible bundle. Sellers bundle the product with extras and then mention the bundle only in the title or bullets. The main image still shows the bare core product. You’ve added COGS, freight, and packaging complexity without buying a single extra click. Bundling without main-image visibility is half a strategy.
Empty slots. Three clean product shots and four unused slots is a failing gallery no matter how nice those three shots are. The slots are free. Fill them.
Infographic text nobody can read. Eight features in 10-point type, designed on a 27-inch monitor, viewed on a phone. If you can’t read the callout at thumbnail size, it isn’t in the image as far as your buyer is concerned.
Seven shots of the same thing. Rotating the product 45 degrees seven times is not a gallery. Each slot needs a different job, which is exactly what the one-line-per-slot brief prevents.
A gallery that outruns the product. This is the expensive one. Beautiful photography on a generic item buys you a temporary lift and a slow decline, the way it did on that wood therapy launch. Differentiation has to be baked in before you source, not bolted on after.
Ignoring the gallery once the listing is live. Image order is a variable you can test, and most sellers set it once and never touch it again. Test it weekly in month one while the listing is still forming its signals.
Should you shoot this yourself or hand it off
If you’re on your first product and you have time before inventory lands, do it yourself. A sample, a window, a white sweep, a phone camera, and a free editing tool will get you a compliant main image and a usable gallery. Write the seven-line brief first. The brief is worth more than the camera.
Hand it off when the cost of a slow launch exceeds the cost of the work. If you’re paying storage on inventory that’s sitting because your listing doesn’t convert, a $300 gallery rebuild pays for itself in days. The clients who hand us the whole launch get the gallery built off samples while production runs, so the listing is complete on the day stock checks in. That’s what our done-with-you clients are buying, and you can book a discovery call if you want us to look at your current gallery before you spend anything.
Photography is also one input of nine. If the gallery is fixed and the listing still underperforms, work through the full 9-point listing audit we run on every client, and if you’re Brand Registered, the 5-module A+ Content layout is the next place conversion hides.
Frequently Asked Questions
How many images does Amazon product photography need to cover?
Seven, if you’re a standard seller. Six images plus one video if you’re Brand Registered. Fill every slot you’re entitled to, because an unused slot is a sales asset you’re giving away for free.
Can I use AI-generated images on my Amazon listing?
For gallery and lifestyle images, yes, as long as the image accurately represents the real product. For the main image, AI is acceptable only as cleanup or compositing of a genuine product shot, never as invention. For Brand Registry or GTIN Exemption proof photos, no. Amazon wants real photos of real packaging with your brand physically printed on it.
What are Amazon’s main image requirements?
Pure white background at RGB 255, 255, 255. The real product filling about 85% of the frame. No text, logos, badges, borders, or added graphics. At least 1,000 pixels on the longest side so zoom activates, with 1,600 pixels or more recommended. Check Seller Central for the live guidelines before shooting, since Amazon updates them.
How much does Amazon product photography cost?
On our own client launches, a full photographer plus listing content package ran about $800 before AI tools were usable. An AI-assisted version of the same workflow, briefed properly, now lands closer to $200 to $400. Doing it yourself off a sample costs the price of the sample, which for us is around $150 per supplier.
Do I need a video on my Amazon listing?
Only if you’re Brand Registered, and only if the product needs demonstrating. Amazon’s published figure is that video can lift conversion by up to 10%. If “how does it work” is your buyer’s blocker, take the trade. If the product’s use is obvious, a seventh still image is usually worth more.
How often should I update my listing images?
Weekly for the first month of a launch, testing image order and callout wording across four cycles. Quarterly after that. Once a listing is ranking and converting, frequent changes add noise and risk breaking what’s working.
The Bottom Line
Seven slots, seven different jobs, and a one-line brief for each before anyone picks up a camera. That’s the whole system. The main image carries your differentiation into the search grid, the bundle shot confirms it, the scale shot prevents a return, the infographic and lifestyle shots answer “what is it” and “is it for me”, and the objection killer removes whatever is stopping the sale. Get those right and you’ll convert traffic you’re already paying for. Get the product right first, though, because no gallery has ever rescued a generic item from a price war. We build this gallery for every client launch we run, off samples, before inventory lands.
