Jewellery Virtual Try-On: A Practical Guide for Jewellery Brands and Sellers

Jewellery virtual try-on lets a shopper see a necklace, a pair of jhumkas or a full bridal set on a photo of themselves before they buy. Done properly, it answers the two questions a packshot never can: where does this sit on me, and how big is it really. This guide explains how AI jewellery try-on works, what separates a convincing result from a pasted-on one, and how to roll it out across your catalogue without a new shoot.

Why jewellery is hard to sell from a packshot

Clothing has size charts and model measurements. Jewellery has almost nothing that translates to the body. A product photo on a white background or a velvet bust leaves the shopper guessing about the things that decide the purchase:

• Length and fall. Does the haram rest at the collarbone or drop to the sternum? Will the pendant sit above or below a neckline?
• Scale. A jhumka that fills the frame in a macro shot may be delicate in real life, or far more dramatic than the shopper expects.
• The set as a whole. Bridal and festive jewellery is bought as a look. Seeing a necklace, earrings and a maang tikka in three separate photos is not the same as seeing them together on a face.
• Skin tone and styling. Gold, kundan, polki and silver each read differently against different complexions and outfits.

Shoppers fill those gaps with guesswork, enquiries on WhatsApp or Instagram, and sometimes a return. Try-on replaces the guesswork with a picture.

How AI jewellery try-on works

With the FashClick AI jewellery try-on, the flow is three steps:

1. Bring your product photo. Upload a packshot, or sync your whole catalogue from your store. A clean, well-lit shot of the piece is all it needs.
2. Pick the piece type. Choker, jhumka, kada, maang tikka and so on — or choose Auto and let it identify the piece from the image. The type is what tells the model where the piece is worn and how big it should be.
3. Place it on the photo. The piece lands where it is actually worn, at a true-to-life scale, on the shopper’s own photo or a model photo you have chosen.

Crucially, jewellery try-on is additive. The person’s face, clothes and background stay exactly as they were, and only the jewellery is added. For a shopper, that is the point: it is still their photo, now wearing your piece.

No 3D pipeline. Because the try-on is generated from a 2D product photo, you do not need 3D scans, CAD files or a macro reshoot. Every piece you already photograph can be tried on.

Placement and scale: what separates good from bad

Jewellery is small, and small is where try-on usually falls apart. Pieces float off the skin, get duplicated, or balloon to twice their real size. The fix is to treat every piece type as having its own anatomical rule. FashClick works across eight areas of the body:

Body area	Example piece types	Where it lands
Head & hairline	Maang tikka, matha patti, tiara	Along the centre parting, pendant at the hairline
Ears	Studs, hoops, jhumkas, chandbalis, cuffs, helix stacks	At the lobe or high on the cartilage, matched on both ears
Nose	Nath, nose stud, septum ring	On the nostril the reference shows; no chain invented if the piece has none
Neck	Choker, pendant, mangalsutra, rani haar, layered	Each length where it really falls, from mid-neck to the sternum
Wrists	Bangles, kada, bracelets	Where wrist meets hand; a bangle set is split across both wrists
Hands	Rings, cocktail rings, haath phool	Around the finger with the stone on top
Waist & torso	Waist chain, body chain, brooch	On the hips over clothing; a brooch where the neckline suits it
Ankles	Anklet	Around the ankle, just above the foot

In total that is more than 45 piece types, including nine distinct ear piercing spots, deliberately spanning Western, Indian and South-Asian, Middle Eastern, African and East-Asian jewellery. Scale matters as much as position: a nose stud should stay tiny and a helix cuff compact and high on the ear, sized against the person in the photo rather than stretched to fill the frame.

Bridal and festive sets in one pass

For Indian jewellery sellers in particular, the single piece is rarely the sale — the set is. FashClick can place a complete set in a single generation: a necklace, jhumkas, a maang tikka and a nath together, so the shopper judges the whole look at once.

Two details make sets look right rather than assembled:

• Pairs are matched. Earrings go on both ears, not one.
• Stacks are split sensibly. A bangle set is divided between the two wrists instead of being duplicated onto each.

If you sell ethnic wear alongside jewellery, the same thinking applies to the outfit — our story on how Palki Saree lets shoppers see a piece on themselves shows the clothing side of it.

Keep the look, or restyle around the piece

There are two sensible ways to present a piece, and they suit different jobs:

• Standard try-on keeps everything about the photo and adds only the jewellery. Use it on product pages, where a shopper wants to see the piece on themselves as they are.
• Style My Look flips that round: the outfit and hair change to show the piece off, while the person’s identity is preserved. A rani haar might get a lehenga and an updo; a delicate pendant a V-neck and open hair. Use it for lookbooks, festive campaigns and social posts.

AI try-on or AR for jewellery?

AR try-on, the live camera overlay, has long been associated with rigid accessories, and it does let a shopper turn their head and watch an earring move. The catch is that every piece needs a 3D model. AI try-on trades the live view for a photorealistic still, works from the photos you already have, and can place an entire bridal set in a single image. For a catalogue with hundreds of designs and frequent new arrivals, that usually decides it. Our comparison of AI and AR try-on goes deeper on the trade-offs.

Getting your packshots ready

Good input is the biggest lever on output quality. Before you switch try-on on, check your product images against this list:

• The piece is sharp and fully in frame — no cropped chains or clipped drops.
• Lighting is even, so stones and metal finish read clearly rather than blowing out.
• The background is plain and the piece is not tangled with props.
• For sets, every component is visible so nothing has to be guessed.
• Piece types are set correctly in your catalogue, or left on Auto where you are unsure. A pendant tagged as a choker will be placed like a choker.

Rolling it out on your store

Shopify

Connect your store, sync your catalogue and enable a “Try it on” button on jewellery product pages. It is the same flow as clothing try-on, so if you already use FashClick for apparel there is nothing new to learn. The FashClick Shopify app page covers setup.

Custom storefronts and marketplaces

If you run your own storefront or a multi-seller marketplace, call the API directly. Jewellery is uploaded under the Accessories category, and the request and webhook flow is documented in the developer docs.

A sensible first rollout

1. Start with the categories where placement causes the most doubt: long necklaces, statement earrings and bridal sets.
2. Test your hardest pieces first — layered necklaces, ear stacks, a nath with a chain.
3. Tell shoppers clearly what happens to the photo they upload, and keep that promise.
4. Reuse the best results as on-model imagery for listings and social posts.

The bottom line

Jewellery sells on how it looks worn, and a packshot can only hint at that. AI jewellery try-on shows the piece at its real size, exactly where it sits, on the shopper’s own photo — and it does it from the images you already have. See how FashClick jewellery try-on works, then use the 5 free credits on signup to test it on your own hardest pieces before committing to a plan.