Product Photography Cost Per SKU: The Number That Runs Your Catalog
Ecommerce teams talk about photoshoot budgets. Catalog operations run on a different number: what it costs to get one product photographed, approved, and live. That number is the one that scales with the business, and most brands have never calculated it honestly.
The short version: cost per SKU is roughly flat while revenue per SKU falls off a cliff. That single fact explains why the bottom 60% of almost every catalog is presented badly.
Calculate it properly
Most teams compute cost per image, which flatters the number, or divide the photographer's invoice by the SKU count, which understates it badly. The honest formula:
Total loaded cost ÷ SKUs that shipped usable imagery
Total loaded cost includes:
- Crew — photographer, assistants, stylist, hair and makeup
- Studio or location, plus permits
- Sample logistics — pulling, shipping both ways, returns, insurance
- Post-production, per image, at your actual approval standard
- Your team's hours — briefing, attending, selecting, revising, uploading
- Reshoots and rejects
- A share of any usage licensing
Two denominators matter. Use SKUs that shipped, not SKUs photographed — the ones that got dropped still consumed budget. And count a colorway as a SKU, because for imagery purposes it is one.
Run this once on your last production. The result is usually two to four times what people assumed, and it is almost always the loaded team hours that account for the gap.
Why the long tail stays ugly
Here is the structural problem, and it is not a failure of anyone's judgment:
Per-SKU imagery cost is roughly flat. The hundredth product costs about what the tenth did — a bit less through batching, but not fundamentally different.
Expected revenue per SKU is anything but flat. It follows the usual steep curve: a small number of styles carry the business and the tail earns progressively less.
So there is a crossover point where imagery cost exceeds any defensible share of a product's expected margin. Below it, the rational decision is the cheapest possible treatment — a flat lay, a ghost mannequin, or the supplier's own photo.
Which means the products that most need help standing out get the least, and their underperformance then justifies the original decision. The loop is self-sealing, and it is where most catalogs quietly lose money.
What each SKU actually needs
A typical apparel PDP set:
| Image | Purpose |
|---|---|
| Front, on-model or ghost | The primary listing image |
| Back | Fit and construction |
| Detail ×1–2 | Fabric, hardware, trim |
| Lifestyle or in-context | Scale, styling, desire |
| Additional colorways | Accuracy per variant |
Four to eight images per SKU, multiplied by colorways. A 200-style collection in three colorways is 600 variants and, at six images each, up to 3,600 individual assets per season.
That multiplication is the whole reason per-SKU thinking matters. Budgets are approved per shoot; costs are incurred per variant.
What actually moves the number
Batching. More SKUs per setup amortises the fixed cost. Real, and limited by attention — quality degrades late in a long day.
Simplifying the standard. Fewer images per SKU, less retouching. Cuts cost directly and visibly.
Offshoring post. Reduces the per-image retouch line. Adds turnaround time and a review burden that partly offsets it.
Shooting less often. Fewer, bigger productions. Amortises well and hurts speed to listing — new products wait for the next shoot.
Each of these trades something real. None of them changes the shape of the problem, which is that cost per SKU is flat while revenue per SKU is not.
What changes the shape
The reason AI production is interesting here is not that it is cheap per image — it is that it decouples per-SKU cost from crew time.
Once an approved treatment exists, the marginal cost of the next SKU is generation plus review, not another slice of a shoot day. The fixed cost moves up-front into the creative direction, and the marginal cost falls sharply.
That flattens the crossover point. Products that could never justify a shoot day can be presented to the same standard as the hero styles. For a catalog with a long tail — which is most catalogs — this is where the revenue argument actually lives, far more than in the per-image price.
Two honest caveats. First, the up-front direction work is real and should not be skipped; a bad treatment scaled across 600 variants is worse than no treatment. Second, review does not scale to zero — the hard garments still need looking at, per item. Budget for both, and see flat lay to on-model for which garments need the attention.
A model you can fill in
For your own catalog, work out:
- Loaded cost of last season's production (all lines above, including your hours)
- SKUs that shipped usable imagery
- Current cost per SKU = 1 ÷ 2
- Expected gross margin per SKU over its life — for a median SKU, not a hero
- Imagery as a share of margin = 3 ÷ 4
If step five is uncomfortable for the median product, you have quantified why your long tail looks the way it does. Then model the alternative honestly: up-front direction cost, plus per-variant generation and review, divided across the same SKU count.
If you would rather not assemble step two by hand, the photoshoot calculator reads your storefront, counts the products that need imagery, and prices the coverage tiers against that count.
The full side-by-side is in AI versus traditional photoshoot cost, and the traditional line items are broken down in what a fashion photoshoot costs.
Frequently asked questions
How do you calculate product photography cost per SKU?
Take the fully-loaded cost of a production — crew, studio, post-production, sample logistics, and your own team's hours — and divide by the number of SKUs that shipped usable imagery. Not by images produced, and not by SKUs photographed.
How many images does each SKU need?
Most apparel PDPs use somewhere between four and eight images: a front view, a back view, one or two detail shots, and one or more on-model or lifestyle frames. Multiply by colorways, because each colorway is effectively a separate SKU for imagery purposes.
Why is the long tail of a catalog photographed worse?
Because per-SKU cost is roughly flat while expected revenue per SKU falls sharply. Below a certain revenue threshold the imagery investment stops making sense, so those products get the cheapest treatment — which suppresses their performance further.
Do colorways need separate photography?
For accuracy, yes — recolouring an image is a common shortcut and a common source of returns, because fabric renders colour differently than a colour swap suggests. Colorway coverage is where per-SKU costs multiply fastest.
What is a good cost per SKU target?
There is no universal figure — it depends on price point, margin, and expected units. The useful test is relative: your imagery cost per SKU should be a defensible fraction of the gross margin that SKU is expected to generate over its life.
Price your catalog with the photoshoot calculator, or explore Astria for fashion and ecommerce.