Product photography and video: what each marketplace requires

There is a scene that repeats in almost every operation that grows fast. The seller hires a photographer once, at the beginning, for the first batch of products. Twenty references go out, eighty photos come back in a Drive folder, and from then on the catalog is built with whatever is at hand: the photo the supplier sent, a render from the factory catalog, a screenshot of the box, an improvised cutout on a background that is white-ish at best. Two years later there are four hundred listings and nobody knows which image is live on which channel.

That cost never shows up on an invoice. It shows up in CTR — people do not click — and it shows up in conversion — the ones who do click do not buy, because they never understood what they were looking at. And since both marketplaces use conversion to decide who keeps getting shown, a weak gallery does not only cost you today’s orders: it costs you future impressions.

The other problem is duller and just as expensive: technical rejection. A main image with the wrong background, with text baked into it, or with fewer pixels than required does not merely look bad. It can block the listing, disable zoom, or keep you out of the catalog page entirely. Plenty of sellers discover this only after a listing has gone weeks without traffic and nobody can explain why.

This article separates the two things that usually get mixed together: the technical requirements each platform enforces before it accepts your material, and the composition decisions that actually move the commercial needle. The first set is an exam you have to pass. The second is where the click is genuinely won or lost.

the first image does not sell the product: it sells the click

It helps to picture the exact moment your main image does its job. It is not sitting on a detail page with room to breathe and context around it. It is in a grid of results, on a phone, competing with twenty other thumbnails of identical size, while a thumb scrolls faster than anyone can read.

In that context the main image does not have to explain the product. It has to accomplish one thing: stop the thumb and let the shopper recognize, in half a second, that this is what they were looking for. Everything else — materials, dimensions, what is in the box, how it is used — gets resolved in the following images, once the click is already yours.

Two opposite mistakes follow from that. The first is a main image that is too empty: the product marooned in the middle of a vast white canvas, so small at thumbnail size that it is indistinguishable from the competition. The second is a main image that is too loaded: the product with every accessory, the box, the manual and three overlapping views, which at thumbnail size collapses into an illegible smudge.

The practical rule that works on both platforms is simple: open it on your phone, at thumbnail size, next to your competitors’. If it is not readable at that scale, the megapixel count does not matter. That is the only test that reproduces the real moment of decision.

Amazon’s technical requirements

Amazon draws a hard line between the main image and the rest of the gallery, and it is the main image where the rules bite.

Amazon’s style guide asks the main image to sit on a pure white background — RGB 255, 255, 255 — with no heavy studio shadow and no near-grays that merely read as white. An off-white, a light gray or a gradient background is one of the most common causes of rejection, and the rejection is not always immediate: the listing lives quietly until an audit flags it.

The product should fill most of the frame — the guide talks about roughly 85% or more — and appear whole, not cropped. The main image allows no text, no overlaid logos, no watermarks, no “free shipping” badges and no inset images. A logo physically printed on the packaging is fine, because it is part of the product; a logo you added in the editor is not.

On resolution, the criterion that matters is zoom. Amazon enables zoom once the longest side of the image clears a threshold, and the platform’s own references point to 1,600 pixels as the number where zoom behaves properly, with a recommendation to go higher when you can. There is a lower technical minimum the system will accept, but staying there means publishing without zoom — a feature shoppers reach for precisely when they are about to decide. The usual formats are JPEG, PNG, TIFF and GIF, and JPEG is what almost everyone ends up using for the balance between quality and file weight.

Secondary images get far more freedom: lifestyle backgrounds, overlaid text, infographics, comparisons and models using the product. That is where the persuasion work actually happens.

Detail page video is a separate story. Uploading through Upload and Manage Videos in Seller Central is reserved for sellers enrolled in Brand Registry, and every video goes through review before it goes live. The rejection reasons that repeat most often are mentioning prices or promotions, referencing other sellers or marketplaces, and showing something that does not match the exact product on that page.

Glossary: unified catalog, one real product and many listings →

MercadoLibre’s technical requirements

MercadoLibre came to the white-background rule later than Amazon, but today a pure white main photo (#FFFFFF) is required in most categories, and the platform has been tightening the rule category by category, with specific guides for apparel, footwear, accessories, bundles and several more.

On size, Meli’s public recommendation hovers around 1,200 × 1,200 pixels as the working standard, with a lower technical minimum that is best avoided: in competitive categories an image barely above the floor is visible. Accepted formats are JPG, JPEG and PNG, with a per-file weight cap documented in the platform’s help center.

Two details surprise anyone coming from Amazon. First: the maximum number of photos depends on the category, and on listings with variants the count is handled per variant, not per listing. Second: Meli’s category guides are far more prescriptive than you expect — they specify the angle a garment must be shot from and whether the product may be photographed on a mannequin or laid flat. Those are not style suggestions; they are approval requirements.

Videos on MercadoLibre follow their own logic, closer to social media than to a product page. The format is vertical (9:16), with a duration that runs from roughly ten seconds to a minute, and the platform expects voice or music — if there is none, Meli adds a royalty-free track so the video can be approved. There are safe zones to respect: no text where the product card or the interface buttons sit. Review takes a couple of business days and costs nothing.

All of the above changes often and varies by category. Before you rebuild an entire gallery, open the photo guide for your category in MercadoLibre’s seller help center and the style guide for your category in Seller Central. What we describe here is the general framework; the number that applies to you lives in the official source.

the four types of shot that do the work

Past the technical part, the interesting question is which photos to take. After looking at hundreds of galleries, the shots that resolve real objections come down to four, and almost every weak gallery is weak because one of them is missing.

• Scale. The most frequent doubt and the least addressed: how big is this thing? A hand holding the product, the product next to an everyday object, or a shot with dimensions overlaid. It is the shot that prevents the most returns, and “it was not the size I expected” returns are among the most expensive: you pay the outbound shipping, you pay the return, and sometimes the item does not come back sellable.
• Use. The product in its context, operated by a person. It answers “will this do the thing I actually want to do?”. In home, tool and sports categories it is usually the highest-converting shot in the set.
• Detail. A close-up of whatever justifies the price: the stitching, the material, the connector, the finish, the gauge of the metal. It is the shot that holds your price when you are up against someone cheaper.
• Comparison. The variant, the bundle, or what is included. If you sell sizes, colors or packs, this shot keeps the buyer from choosing wrong — and a wrong choice ends in a return or a two-star review that was never the product’s fault.

You can build infographics on top of those four, but watch one thing: overlaid text either reads on mobile or it does not read at all. Three large words work; a nine-point paragraph is decoration.

how to know whether a gallery change worked

Here is the part almost nobody does, and it is what turns photography into a business decision instead of a matter of taste.

An image change affects two different metrics and they have to be read separately. If the main image improved, what should move is click-through rate: the same impressions producing more visits. If the rest of the gallery improved, what should move is conversion: the same visits producing more orders. Change both at once and you will not know which one worked.

Hence the minimum discipline: one change at a time, with the date written down and an honest comparison window. Honest means comparing against a period equivalent in traffic and seasonality, not against Buen Fin week. And it means watching margin too: a gallery that lifts conversion but shipped alongside a 10% discount proved nothing about the photos.

It is also worth measuring what the platform dashboard does not show: the return rate per SKU. A gallery that explains scale and package contents properly lowers returns, and that reduction is real money even if conversion does not move a point.

Glossary: Buy Box, the buy box that decides who makes the sale →

the invisible work: one product, different galleries per channel

When you sell the same product on Amazon, on MercadoLibre and on your own store, the galleries drift apart on their own. You swap the main image on Amazon because it got rejected, you add two lifestyle shots on Meli for a campaign, and Shopify is still running the supplier’s originals. Nobody decided that; it simply happened.

The drift costs you twice. First, a shopper comparing you across channels sees what look like different products. Second, when you want to evaluate whether the new gallery worked, you cannot: each channel has a different version and there is no clean before and after.

Holding this together is not a matter of more people staring at screens. It is a matter of keeping an inventory of assets per product: which shots exist, which main image is the approved one, which version is live on each channel, and since when. Without that record, every gallery audit starts from zero.

how this looks in iqseller

Worth saying plainly: iqseller does not edit your photos, does not write copy and does not upload video. It is not a content tool and does not pretend to be one.

What it does contribute is the structure that makes a content change measurable. The Catalog module holds the Parent → Model → SKU tree: one real product at the top, its models and variants below, and each channel’s listings hanging from the variant they belong to. With that structure in place it becomes obvious which listings in the same family are misaligned across channels, which ones have an incomplete record, and which ones are orphaned on a variant you no longer sell.

The measurement runs on that same structure. The Profitability module reads net margin per SKU using the COGS you loaded, the commissions from Amazon’s settlement and from MercadoLibre orders, FBA and Full fees, shipping, advertising, VAT and withholdings. So when you rebuild the gallery for an entire family, you can look at conversion and net margin for the same SKU before and after, instead of trusting the feeling that “it looks better now”. Alerts tell you when the move went in the wrong direction.

The panel does not tell you which photo to take. It tells you whether the one you took worked.

where to start without redoing everything

Rebuilding four hundred galleries is not a project; it is a fantasy. The version that actually gets finished is organized by revenue concentration.

Sort your products by sales over the last ninety days and keep the ones that make up most of your revenue. On that short list, do three things: verify the main image meets your category’s requirements on each channel, check that all four shots are present — scale, use, detail, comparison — and write down the date of the change. Leave the rest of the catalog for later, once the criteria are proven.

And schedule a quarterly requirements review. Both platforms adjust their guides frequently, sometimes without a loud announcement, and what is approved today can be flagged in the next audit. Confirming the current requirement in Seller Central and in MercadoLibre’s help center takes fifteen minutes; discovering it through a deactivated listing in the middle of peak season costs a great deal more.