How fashion retailers run virtual try-on inside WhatsApp, the flow step by step, and what W for Woman got: a 4× engagement uplift and 65% WhatsApp engagement.
Fashion retailers offer virtual try-on through WhatsApp by making it a conversation: the shopper opens a chat with the brand, picks a piece, and gets back a generated image of themselves wearing it, with no app or microsite. Whilter ran this for W for Woman with Snype, with a 4× engagement uplift and 65% WhatsApp engagement.
Because a try-on that lives on its own app or microsite forces a detour, and most customers never take it. That’s one of three problems the W for Woman case starts from. A static catalogue photo shows the garment, not the shopper in it, so the buyer is left guessing before they commit. A personal visual for every shopper across a high-SKU catalogue is beyond what a studio can produce by hand. And a separate destination loses the shopper before the try-on even starts.
Putting the try-on in WhatsApp deals with the third problem directly. W for Woman’s version needed no new app and no microsite. It ran in a channel shoppers open daily.
The framing matters as much as the channel. Instead of pushing a static asset at people, the try-on becomes something a shopper asks for in the moment, so each image exists because someone wanted it.
As a short conversation where the shopper asks and the brand answers with an image. The flow Whilter documents for W for Woman has three steps:
There’s no studio in the loop. Each image is model-generated when the shopper asks, so coverage grows with the catalogue instead of with photographer hours. Snype is the Whilter engine behind this. It supports real-time generation through an API as well as scheduled batch runs, and its face-replacement and virtual try-on capability lets a customer see the product on themselves from a single source asset.
The same thread can go further. Whilter’s conversational commerce pattern runs discover, ask and try on, then check out, without leaving WhatsApp or Instagram.
A 4× engagement uplift, an 86% image-generation success rate and 65% WhatsApp engagement. W for Woman is a leading Indian fashion brand, and its case page ties each number to a cause.
The 4× is attributed to the imagery being personal: a personalised image is a stronger draw than a static product shot. The 86% is the success rate of generation across a high-SKU catalogue, achieved without one-to-one manual work. And the 65% is credited to delivering the try-on inside WhatsApp instead of a separate app.
Those are engagement figures, and engagement isn’t revenue. If you run this, agree up front what counts as a win for you, whether that’s orders that start in the thread, fewer returns, or something else you can pull from your own data.
Product data, a proper WhatsApp channel, a brand look written down, and a plan for the image that doesn’t come back.
Start with the catalogue. A high-SKU catalogue is the reason W for Woman couldn’t shoot this by hand, and it’s also what the try-on draws from, so clean product images and consistent SKU naming are worth fixing first. Whilter’s commerce integrations bring in live catalogue, pricing and stock, and Snype takes first-party data from a CRM, CDP, spreadsheet or product database, with you choosing the fields.
Next, the channel. Whilter runs on the Meta-Verified WhatsApp Business API, which is what gives a thread the green-tick brand trust. Plan for the official Business API on your side too, rather than improvising on a personal number.
Then the look. Snype’s brand-safe templates enforce your look and tone, but somebody has to decide what that look is: framing, background, styling. Write it down before the first image goes out.
Last, the failure case. Decide what the shopper sees when an image doesn’t come back, because no generation system returns a good one every time. A catalogue photo plus an offer to bring in a person is a sane fallback. Konne CX, Whilter’s conversation platform, hands high-value moments to a human agent who carries the full context, so the customer never repeats themselves.
Treat the photo as personal data from the first message: ask before you take it, say what it’s for, and decide up front how long you’ll keep it. A try-on only works if the shopper hands over something to work from. WhatsApp makes that easy, which is exactly why the consent line belongs in the chat itself and not behind a terms link.
Don’t reuse the image for anything other than the try-on without a separate yes, and make sure you can honour a deletion request.
On the vendor side, Whilter maintains an ISO 27001, SOC 2 and GDPR security posture and is built to be DPDP-aware. Your first-party data stays yours, and so do the images generated. Konne CX can be deployed managed-shared, as a dedicated tenant, on-premise or under your own cloud, so data residency can follow your policy. Ask any vendor, us included, to put those answers in writing before shopper photos go near it.
Snype is the engine. The W for Woman case credits it with running the whole loop as a WhatsApp conversation. If you want the conversation to carry on past the image, with AI and human agents sharing one context on the Meta-Verified WhatsApp Business API, that’s Konne CX. Whilter plugs into your existing stack rather than replacing it.
For the wider retail picture, including whether Whilter does virtual try-on for fashion D2C (it does, and W for Woman is the proof), see Whilter for retail and D2C.
If you want to scope a WhatsApp try-on against your own catalogue, get in touch.
Published 2026-10-06 · Whilter.AI
