The three-way constraint: fit, policy, and ad approval
Before touching a light, understand what each destination will and won't accept. The same image set rarely serves all three without adjustment.
| Destination | Key constraint | Practical implication |
|---|---|---|
| Amazon main image | Human model required for clothing; pure white background | On-model is mandatory, not optional |
| Amazon secondary images | No sexually explicit or obscene content, at Amazon's discretion | Suggestive posing is the usual suppression trigger |
| Meta ads | Allowed if it doesn't violate Adult Nudity policy | Close crops on the body read as violations |
| Your own store | No external policy | Where your most detailed imagery belongs |
| Pinterest / Google Shopping | Conservative automated filtering | Favor flat lay and product-focused crops |
The workable strategy is one shoot, three edits. Capture a full set on-model, then derive a conservative crop for marketplaces and ads, and reserve the fuller set for your own product detail page where you control the rules. Trying to shoot separate sessions for each destination is how intimates brands burn their photography budget.
Amazon rarely tells you which specific image triggered a suppression, and the review queue for reinstatement can run several days. For a seasonal intimates drop, that's the launch window. Shoot conservative for the marketplace set from the start rather than testing the boundary.
What shoppers actually check before buying intimates
Intimates has among the highest return rates in apparel, and returns in this category are overwhelmingly fit-driven rather than taste-driven. Your image set's job is to pre-answer the fit questions, because a shopper who can't resolve them either doesn't buy or buys three sizes.
The questions, roughly in the order shoppers ask them:
- How does the band or waistband sit? Needs a straight-on shot at natural tension, not stretched flat.
- How much coverage is there, front and back? Both views required. Missing back views are the single most common gap in intimates listings.
- Is the fabric opaque? Sheerness is a fit-and-confidence question. It must be visible in at least one image, honestly.
- What is the closure, strap, and adjuster hardware? One detail macro covers all three.
- What does this look like on a body like mine? Multiple model sizes, if you can afford them.
- What is the actual color? Nude and blush tones shift dramatically under warm light — this is where colour accuracy stops being a nicety.
Lighting and technique for delicate fabrics
Lace, mesh, silk, and microfiber each fail differently under bad light. The common thread is that intimates fabrics are either semi-transparent, highly specular, or both — and the default e-commerce lighting setup handles neither well.
Lace and mesh need directional light raking across the surface to read as texture rather than grey mush. A large softbox as key at roughly 45 degrees, with a second light kicking from behind and slightly to the side, separates the pattern from whatever is underneath it. Flat frontal light is the enemy here: it flattens lace into a smear.
Silk and satin are specular. Broad, soft sources are essential — a large softbox or a scrim panel — because small hard sources produce blown highlights that read as white blotches and destroy the perception of colour. If you're seeing clipped highlights on a satin cup or a slip hem, your source is too small or too close.
Nude, blush, and skin-adjacent tones are the category's colour-accuracy nightmare. These sit close enough to skin tone that any warm cast makes the garment disappear against the model, and any cool cast turns it grey. Shoot a colour checker in the first frame of every setup and correct against it. Eyeballing nude tones on an uncalibrated monitor is how brands end up with a return reason of "colour not as pictured."
Intimates fabrics crease from the packaging and hold the crease under studio heat. Steam before the model dresses, then again between setups. Wrinkle removal in post is possible but on lace and mesh it costs texture detail — it is materially cheaper to fix on set.
On-model, mannequin, and flat lay: when each applies
Amazon requires a human model on the main image for women's and men's clothing, which settles the question for the marketplace hero shot. Everything else is a judgment call, and each approach has a distinct role.
Flat lay & ghost mannequin
- Cheapest to produce and to reshoot
- Zero policy risk on any platform
- Best for showing construction and hardware
- Poor at answering fit questions
- Cannot serve as an Amazon clothing main image
On-model
- Required for Amazon apparel main images
- Answers fit, drape, and coverage directly
- Highest conversion impact in the category
- Most expensive; casting and scheduling heavy
- Carries all of the policy risk
The practical mix for most intimates brands: on-model for the hero and two to three supporting views, ghost mannequin or flat lay for the construction detail and colourway variants, and a macro for fabric. Ghost mannequin earns its place here specifically because it shows a garment's three-dimensional shape without a body in frame — which sidesteps ad-review risk entirely while still communicating more than a flat lay.
Colourway variants are the other budget lever. Shooting every colour on-model multiplies your day rate by the number of colours for imagery that differs only in hue. Most brands shoot one colourway on-model and generate the remainder — either via careful recolouring or an AI retouching pass — which is exactly the kind of mechanical, high-volume work tools like Retouchable are built for. The caveat: recolouring nude and blush tones needs verification against a physical sample, because these are precisely the shades where an approximate match becomes a return.
Working with AI models in a sensitive category
AI-generated models solve real problems for intimates brands — size range coverage without casting six fit models, regional variation, and imagery for colourways that never justified a shoot. They also carry category-specific risks worth naming plainly.
- Disclosure expectations are higher here. Fit is the purchase decision. A shopper who discovers the body in the image was generated has reason to distrust the fit signal, which is more damaging in intimates than in, say, outerwear. Label AI imagery, and keep at least one verifiably real fit reference in the set.
- Anatomical artifacts are more visible. Straps, band tension, closure hardware, and the way a garment meets the body are exactly where current generation models are weakest. Every generated intimates image needs a human review pass focused on strap continuity and hardware.
- Policy review does not distinguish. An AI-generated image that reads as suggestive gets rejected on the same terms as a photographed one. Generation does not buy you latitude.
- Regulatory labeling applies. The EU AI Act's transparency obligations cover synthetic imagery in commercial contexts, and marketplaces are building their own disclosure fields. Build labeling into your upload workflow now rather than retrofitting.
Illustrative split for a mid-size intimates brand. Note where the money actually sits: casting and day rates dominate, which is why colourway multiplication is so expensive and why brands look to reduce shoot count before they look at retouching cost.
A shot list you can hand to a photographer
Standardizing the shot list is what makes an intimates catalog look coherent and what keeps you from discovering a missing back view three days after launch.
- Hero, on-model, front, straight-on. Pure white background for the Amazon main image. Neutral standing pose, arms clear of the garment, full band and strap visible.
- Back view, on-model. Non-negotiable. Shows closure, strap configuration, and rear coverage.
- Three-quarter, on-model. Communicates drape and side profile — the view that answers "how does this actually sit."
- Second model size or body type. Same pose, same lighting, so the comparison is legible.
- Fabric macro. Lace pattern, weave, or sheen at close range. Raking light. This one image carries the quality perception for the whole listing.
- Hardware detail. Clasp, adjuster, underwire casing. Shot on a flat or ghost mannequin, not on the model.
- Flat lay or ghost mannequin, full garment. Construction reference and your safest asset for ad creative.
- Scale or fit reference. A size chart overlay or a measured flat, depending on category norms.
Leave enough headroom and margin in the on-model frames that you can derive a marketplace-safe crop without re-shooting. Cropping tighter later is easy; inventing frame you never captured is not.