How A Premium Fashion Brand Cut Photoshoot Costs By 78%
How does a fashion brand use AI virtual try-on to reduce production costs?
A mid-market D2C fashion brand replaced 80% of its catalog photoshoots with AI virtual try-on, cutting annual photoshoot spend from approximately $310,000 to $68,000 — a 78% reduction. The team kept human-photographer shoots for editorial campaigns and seasonal hero imagery, and moved all product-detail-page (PDP) variations, colorways, and lookbook drafts to Drape Try-On. Garment flat-lay photography was already required for inventory, so the marginal cost of producing on-model imagery dropped to roughly $0.075 per look.
- Replaced 80% of catalog photoshoots with AI try-on.
- Annual photoshoot spend dropped from $310k to $68k.
- Editorial / hero photography retained for brand campaigns.
- Time-to-market for a new drop: 6 weeks → 9 days.
- Same-season testing of colorways became economically viable.
Metrics
The challenge
The brand operated a quarterly drop cycle with roughly 60 SKUs per drop. Each drop required a 3-day on-model photoshoot at $25,000–$35,000 including studio, photographer, model, stylist, hair, makeup, and post-production. Across four drops per year plus retouching and reshoots for hero pieces, the total annual photoshoot spend reached approximately $310,000.
The constraint was not budget alone — it was speed. Shooting a colorway required physically having the colorway sample, which sat at the end of a 6-week production cycle. By the time photos were ready, the season was already 40% over.
The approach
The team split the catalog into two tiers. Tier A (the 15% of SKUs driving the season’s marketing story) continued with traditional photography — photographer, model, location, art direction.
Tier B (the remaining 85% of catalog — colorway variants, sizes, simple lookbook entries) moved to Drape Try-On. The team produced flat-lay garment photography in-house (which they did already for inventory) and ran each garment through Drape against a small library of 8 model reference photos covering different body types, skin tones, and heights.
The Pro plan ($39/month) provided the REST API used by their internal catalog tool to batch-generate variations automatically when a new garment was uploaded to the inventory system.
Implementation details
The internal catalog tool was extended in one engineering sprint. When a garment record was saved, a webhook called Drape’s Pro API with each model reference photo as a person input. Results came back in roughly 12 seconds per generation. A simple approval queue showed the marketing team each output for one-click approve or regenerate.
For blazers and structured outerwear (the categories where AI try-on can still produce artifacts), the team kept human photography. This category-level routing was the single most important design decision — it kept brand quality high while capturing the cost saving on the long tail.
Results after one year
Total photoshoot spend dropped to roughly $68,000 (a 78% reduction). Drape Try-On API usage averaged 2,800 generations per month at a cost of around $210 — a small fraction of the savings.
The more interesting result was speed. New colorway variants started appearing on PDPs within days of being approved by the design team, rather than waiting for the next quarterly shoot. The brand began A/B testing color choices mid-season for the first time, which separately drove a measurable lift in sell-through on previously-slow colorways.
Frequently asked questions
Is the customer real?+
Does AI try-on work for all garment categories?+
How long does the integration take?+
Drape Editorial is the in-house research team behind Drape Try-On. We test virtual try-on models against real garment photography weekly and publish what we learn.