Photography · Ecommerce · Sydney
Ecommerce photography in Sydney
A listing image has one job: make the product unmistakable. We shoot hero on-white packshots and macro detail for Sydney ecommerce brands, framed and sized to the specification Google actually publishes, so nothing in the background competes with what you are selling.
Grab the snowball. Give it a spin, and watch it grow as you scroll. Just like your image library.
The short answer
What is ecommerce photography, and what does it cost?
Ecommerce photography is the pair of frames every listing needs: a hero on-white packshot that makes the product unmistakable in a grid, and macro detail that settles the quality question. In Sydney it runs from about $25 to $40 an image at volume, or $1,200 to $1,600 for a catalogue day covering 15 to 25 products.
The difference between a listing that converts and one that gets scrolled past is rarely artistry. It is whether the product is clean, correctly framed and large enough in the frame, which is exactly what the platforms measure and what an AI answer needs in order to identify what it is looking at.
We shoot in our Sydney studio for brands across greater Sydney and New South Wales, and every file is delivered to the specification the platform actually publishes rather than to a house style. Because our Snowball SEO platform automates the search heavy lifting other agencies bill by the hour, more of your budget goes into the imagery. It sits inside our wider Sydney photography service.
The specification
What Google Merchant Center actually requires
Most pages on this subject repeat a version of the rules that was true a few years ago. These are the current ones, read from Google's own documentation on 30 July 2026, with requirements separated from recommendations, because breaking a requirement gets a product disapproved and missing a recommendation only costs performance.
| Element | What Google says | How we shoot it |
|---|---|---|
| Image size | RequiredAt least 500 x 500 pixels, across every category and marketing method. | 1500 x 1500 or above for best performance in all listing formats. |
| File limits | RequiredNo larger than 64 megapixels, and no file over 16MB. | Submit the largest, highest quality original you have, rather than an upscaled copy. |
| Framing | Best practiceThe whole product must be shown accurately, with minimal or no staging. | The product should fill no less than 75 per cent and no more than 90 per cent of the frame. |
| Background | Best practiceNothing that competes with or covers the product. This is where most feeds fall down. | Solid white or transparent is recommended, not required. Staged images that clearly show the product are also acceptable. Avoid transparent on pale products, which can render on black. |
| Overlays | RequiredNo promotional elements at all: calls to action, prices, free shipping or warranty notes, words like best or cheap, condition claims, watermarks, barcodes, or brand and retailer logos unless they are part of the product itself. | Keep the sell in the listing copy, where it belongs, and out of the pixels. |
| Borders | RequiredNo border around the image. | Borders stop the image sitting cleanly inside varied layouts. |
| What it cannot be | RequiredNot a placeholder, not a generic illustration, not a logo, not a single block of colour, and not a product you do not sell alongside the main one. | Additional views belong in the additional image slots rather than crowded into the main one. |
Sources: Google Merchant Center Help, image link [image_link] and the 2026 product data specification update, both read 30 July 2026. Requirements change, so treat any page quoting them, including this one, as current at its date rather than permanent.
The formats
The two shots that do the work, and the four that support them
The hero on-white earns the click and the macro detail closes the doubt. Everything else exists to remove a reason not to buy.
Hero on-white
The listing thumbnail. Product square to camera on a clean white or transparent field, centred, true to colour, filling the frame properly.
Macro detail
Close enough to settle the quality question: weave, grain, stitching, finish, the machined edge. The frame that answers what copy cannot.
Cut-out and clipping
Hand-finished paths so the product sits cleanly on any background, including the transparent version your platform may need.
Multi-angle sets
Front, three-quarter, back and base, sized consistently so a gallery steps through without jumping.
Scale and dimension
A reference frame that makes size obvious without a tape measure, which is the single biggest cause of avoidable returns.
Platform-ready exports
Every crop and file type your store, marketplaces and feed need, named to your product codes.
Worth knowing
Why a real photograph is safer in a product feed
An AI image has to declare itself
Google requires product images created with generative AI to keep their IPTC DigitalSourceType metadata, the tag that marks an image as algorithmically made, and states plainly that you should not remove it. So a generated packshot travels through your feed carrying a label that says it was generated.
A photograph of the actual object carries no such tag, because there is nothing to declare. That is not a marketing line, it is the practical difference between the two routes, and it is worth knowing before you build a catalogue on generated imagery.
There is a second reason, and it is the one that matters more each month. An AI shopping surface has to work out what a product is from the picture as well as the words. A clean, isolated, well framed object is unambiguous. A cluttered frame is a guess, and a guess is a reason to show a competitor instead.
What that means for your shot list
Shoot the real thing, isolate it properly, and keep the frame free of anything you are not selling. That is the whole discipline, and it happens to satisfy the platform rules, the conversion argument and the machine-readability argument at the same time.
It is also why we hand-finish cut-out paths rather than accepting an automatic mask. An approximate edge is visible the moment the product sits on a coloured background, and a soft edge like hair, mesh or glass is where automatic tools give up.
If you also want the product shown in use, in context and with people, that is a different brief and a different frame. It belongs in our brand and lifestyle work, alongside the clean listing set rather than instead of it.
The cost
What a catalogue costs to shoot
On-white and detail work is the most repeatable photography there is, so it should get cheaper per image as the catalogue grows. If it does not, the pricing is not reflecting the work.
Starter set
$400 to $700
Up to about eight products, hero plus one detail each, delivered on white and as transparent cut-outs.
Catalogue day
$1,200 to $1,600
15 to 25 products in a day with consistent lighting and framing across the set. The best value per image.
Rolling programme
Retainer
New lines shot as they land, matched to the existing catalogue so the store never has a visible before and after.
Indicative Sydney ranges, not fixed quotes. The number of products, frames per product and cut-out complexity move it, and we confirm after a free consult.
The brief
What to ask before you book a packshot photographer
Four questions, and the first one sorts most of the field.
What size and framing do you deliver to?
If the answer is not a pixel dimension and a frame-fill percentage, they are shooting to taste rather than to the specification your feed is judged against.
Are cut-outs hand-finished?
Ask to see a cut-out product on a mid-tone background. That is where an automatic mask falls apart and a hand-finished path holds.
Can you match our existing listings?
A catalogue built over years only looks coherent if new work matches the old. Ask whether the setup is recorded so it can be repeated.
What is the rate at volume?
The fortieth product should not cost what the fifth cost. Ask where the rate steps down and at what quantity.
We are happy to answer all four before you commit. If you want the wider picture on shoot types and what they cost, see our guide to product photography costs, or the ecommerce SEO guide for how listings get found in the first place.
Start here
Ready to shoot your catalogue?
Send us the product list and where the images need to go, and we will come back with a shot list, the delivery specification and a fixed price.
Get your free shoot consultGood questions
Ecommerce photography FAQs
Does Google require a white background for product images?
No, and this is the most common misunderstanding. Google recommends a solid white or transparent background because it sits cleanly inside varied layouts, and states that staged images which clearly show the product are also acceptable. What is actually prohibited is anything that competes with or covers the product. White is the safest default rather than the rule.
What size should ecommerce product images be?
At least 500 by 500 pixels, which becomes enforced from 31 January 2027 with warnings running since April 2026. For performance across every listing format Google recommends 1500 by 1500 or larger, capped at 64 megapixels and 16MB. We shoot and deliver well above the minimum so the files survive the next specification change too.
How much of the image should the product fill?
Between 75 and 90 per cent. Below that the product reads as small beside competitors in a grid. Above it, you risk Google's automatic crop cutting into the product, since Google may crop images to focus on the product.
Why does clean product photography matter for AI search?
Because an AI surface has to decide what a product is from the image as well as the text. A clean, isolated, correctly framed product is unambiguous. A busy frame with props, overlays or several items forces a guess, and a guess is a reason to show something else instead.
Can we just use AI to remove the background?
You can, and Google offers tools that do it, but there is a catch worth knowing. Google requires AI-generated product images to keep their IPTC DigitalSourceType metadata and says explicitly not to remove it, so an AI-made image declares itself as AI-made. A real photograph carries no such tag.
Do you supply cut-outs and transparent files?
Yes. Hand-finished clipping paths rather than an automatic mask, which matters on hair, glass, mesh and anything with a soft edge. You get the white version and the transparent version, plus any platform-specific crop you need.
Can you match our existing catalogue?
Yes. Send a handful of current listing images and we will match the lighting, angle, framing and colour so new products sit beside old ones without a visible join, and we record the setup so later additions match as well.