®Every brand that has priced both eventually asks the same question, and gets the same unhelpful answer from whoever is selling. So here is the honest version, which depends almost entirely on how many images you need and how often the product changes.
The first render is expensive. The tenth is not.
A photo shoot has a flat-ish cost per session: studio, photographer, samples, retouching. Do it again next quarter and you pay roughly the same again. A render front-loads almost everything into the first image — modelling, UVs, materials, lighting — and then the marginal cost of the next frame collapses. A second angle is a camera move. A second colourway is a material swap. A second flavour in the range is new artwork on the same carton.
That shape is the whole decision. If you need six images of one product, once, photography usually wins. If you need six images of forty SKUs, in three colourways, refreshed when the packaging changes, 3D wins so decisively that the comparison stops being interesting.
Timing beats cost more often than people expect
You cannot photograph a product that does not exist yet. You can render one. Teams that build their assets from CAD during development have listing imagery, packaging visuals and campaign frames ready before the first production run leaves the factory — which means launch content is not gated on samples clearing customs.
The same asset absorbs late changes. A revised label two weeks before launch is a re-render overnight, not a re-shoot with a new sample.
Where photography still wins
- Organic and irregular subjects — food, fresh produce, anything where the appeal is in the imperfection
- Lifestyle and human context, where casting, styling and location carry the image
- Anything trading on authenticity, where a customer being able to sense CGI would cost you more than the image is worth
- One-off needs on a small catalogue, where there is nothing to amortise the build against
The best-run brands stop treating this as a binary. Hard goods, packaging and variants go to 3D; people, food and lifestyle stay on camera; and the two are graded to match so a product page does not visibly switch pipelines halfway down.
The costs nobody quotes you
- Asset preparation. Clean CAD or a proper reference shoot is the input. Bad input is where render budgets actually go.
- Material development. The first translucent bottle, brushed metal or figured wood costs real time; the range that follows inherits it.
- Review rounds. Renders invite pixel-level feedback that a photo shoot never receives. Agree the rounds up front.
- Storage and versioning. Source scenes are large and become a dependency. Decide who holds them before you need them.
A useful test: count the images you will need over the next two years, not this campaign. If that number is large or the product range is still moving, build the asset. If it is small and fixed, book the studio.
What to ask a supplier
- Do I own the source scenes, or only the exported frames?
- What does an added SKU or colourway cost once the range is built?
- Can the same asset produce AR, turntables and video, or is it stills only?
- How is artwork applied — projected onto the form, or flattened onto a plane?
The last one matters more than it sounds. Artwork projected onto real geometry follows the draft and radius of the pack and can be swapped in minutes. Artwork baked flat means every variant is a rebuild, and the economics that made 3D attractive quietly disappear.