An app called Style Sketch appeared in the app stores of Indonesia, Singapore and Thailand this week. The listing names Meta Platforms Inc. as the developer. There was no blog post, no launch event and no explanation of why those three countries.
The app was found on Oct. 29 rather than announced on it, the way most of Meta’s experiments arrive. It connects to an Instagram account and reads photographs the user has already posted to build a virtual wardrobe. It also accepts pictures of individual garments, either taken with a phone camera or uploaded from a store’s site, and renders the item on the user’s body, adjusting for height, build and skin tone. Two smaller features sit on top: outfit suggestions keyed to the day’s weather, and packing lists assembled for a trip.
Meta did not respond to a request for comment and has not said whether the app will be extended to other markets.
The pattern is familiar inside the company. Threads, the text app that reached 100 million sign-ups in five days in 2023, began as a project inside Instagram and then became its own icon on the home screen. The Meta AI app, relaunched in 2025 with a feed of generated video, followed the same route in reverse, taking a feature that lived inside other products and giving it a store listing. Alongside those successes sit Lasso, a short-video app retired in 2019, and a set of shopping and neighborhood products that never found an audience.
Launching in three middle-income markets first is the cheaper version of that test. Indonesia, Thailand and Singapore have among the highest Instagram penetration rates in the world, phone-first retail habits, and advertising revenue per user well below the American or European average. A product that fails there costs little to wind down and attracts no analyst questions on an earnings call. A product that keeps users opening it daily is worth building out.
The commercial logic runs through the advertising business rather than the app. Retail is Meta’s largest ad category, and since 2024 the company has sold advertisers generative tools that place a product on a model, extend a photograph beyond its frame, and swap backgrounds without a studio shoot. Style Sketch puts a version of that capability in front of the consumer instead of the advertiser, and in doing so collects something Meta has struggled to obtain since Apple’s tracking changes in 2021: a direct statement of what a person owns, what size they wear and what they are considering buying.
That signal has value beyond advertising. A wardrobe assembled from posts is a map of brand affinity over time, and a try-on request is an expression of purchase intent more specific than a like. Meta’s ad system currently infers intent from behavior across its apps; an app that asks users to photograph their clothes obtains it directly, with consent given at the point of upload.
Consent is where the legal exposure sits. Photographs posted to Instagram were uploaded for one purpose, and reprocessing them to build a fitting model is another. The European Union’s rules on data use would make that awkward, and Meta has delayed AI features in Europe before while arguing with regulators about training on user posts. Indonesia’s data protection law, Thailand’s and Singapore’s each impose their own conditions, but none carries the enforcement history of the bloc’s.
The technical problem is unglamorous and hard. Rendering a garment on a body requires the model to preserve fabric, drape, pattern scale and the way a hem falls, and to handle occlusion when an arm crosses a torso. Failures are visible at a glance, which is why virtual try-on has been promised by retailers for a decade and delivered unevenly. Naming skin tone as an input is a partial answer to a documented weakness: image generators have repeatedly rendered darker skin tones poorly, and a shopping tool that does so is not a neutral failure.
The competition is already there. Google added virtual try-on to its shopping results in 2023. Amazon has run a try-on feature for shoes and apparel. Pinterest built body-type filtering into its search. Several startups sell the same capability to brands as software. What none of them has is a social graph that already contains the user’s photographs, which is the only advantage Meta brings to a crowded feature.
A standalone app also buys something a feature inside Instagram cannot: a reason to open a phone on a Tuesday morning when there is nothing to scroll. Feed products compete for attention against each other; a utility competes against the weather app. Meta has said little in public about that distinction, but its product history shows it understands the difference between a feature that adds engagement and a habit that creates a separate one.
None of this is announced strategy, and the company has not described a business model. The plausible routes are a subscription, affiliate commissions on items a user buys after a render, or advertising placed against intent data, and each implies a different set of partnerships with retailers. Meta has tried shopping before, with a Shops product in 2020 and a live-commerce push it shut down in 2022, and neither survived contact with how people actually buy clothes.
The three markets will produce an answer of a sort within a year. If Style Sketch is still listed in Jakarta and Bangkok next autumn, with new features and a wider catalog, it will have earned a bigger launch. If it disappears quietly, the removal will be as unremarked as the arrival, and the experiment will join the list of things Meta shipped small to find out whether anyone wanted them.
