Overview
PerShop offered a personal clothing-store experience built around the brands a shopper already liked. Y Combinator identifies it as a Winter 2019 company founded by Amol Jain and Tony Dong in San Francisco, and now marks it inactive.[1] A contemporary BetaList listing describes following favorite brands, seeing relevant sales and finding products in the shopper’s size.[2]
That is a specific product promise: fewer irrelevant deals, rather than another enormous clothing catalog. The available evidence supports those mechanics. It does not establish why the company stopped, whether customers paid, or whether the business was acquired. Inactivity is an outcome label, not a causal explanation.
Founding Story
YC names Jain and Dong as founders and lists a two-person team. Its current Jain biography identifies him as a WorkWhile co-founder.[1] The record reviewed here does not contain a founder account of PerShop’s origin, division of responsibilities or later transition. A later role is not proof of a pivot or acquisition.
Timeline
- Winter 2019: PerShop participated in YC’s W19 batch.[1]
- May 28, 2019: BetaList featured the brand-following shopping app.[2]
- October 2, 2026: YC lists PerShop as inactive. The linked pershop.app website could not be retrieved during this review. No closure date or transaction was established.[1]
What They Built
The listing connects three choices: favorite brands, relevant promotions and clothing size. A shopper could curate the shopping experience by following brands rather than repeatedly browsing every retailer. BetaList’s sales-and-size description is launch material, not a verified performance result.[2]
The surviving description does not explain how inventory arrived, how size availability was checked, or where a transaction completed. There is no sufficient evidence here for automatic cross-brand checkout, a human stylist, warehouse ownership or a proprietary recommendation model. Those mechanisms should not be added to the company’s history.
Market Position
Target Customers
The documented user was a shopper who wanted deals from selected brands in a relevant size. More detailed demographic or willingness-to-pay claims are unavailable.
Market Size
No reliable company-specific market estimate was located. The existence of shopping tools establishes alternatives, not a reachable market for a small entrant.
Competition
Today, Google advertises price tracking that accepts details such as size and color, then notifies shoppers when a product falls within their budget.[3] A rebuild cannot rely on generic price alerts being absent. That current capability does not explain PerShop’s historical outcome.
Business Model
No observed primary source establishes prices, affiliate contracts, revenue, commissions, gross margin or funding beyond YC participation. Cashback percentages and catalog-scale figures in automated directories are not treated as confirmed. Retailer-funded software is a new hypothesis in the build plan, not PerShop’s known revenue model.
Traction
A dated launch listing shows that the product was publicly presented. It does not establish active shoppers, repeat use, conversion or revenue. No customer cohort or founder-confirmed traction measure was located.
Post-Mortem
The cause of inactivity remains unknown. It would be tempting to blame acquisition costs, inaccurate inventory or retailer fragmentation, but the record does not measure any of them. A plausible operating challenge is keeping promotions relevant to an exact product variant. The launch promise makes that challenge important; it does not prove PerShop failed to solve it.
The rebuild question is therefore narrower than “what new AI fixes the failure?” It is whether a merchant-authorized feed can produce trustworthy, useful size-specific offers for returning shoppers. Shopify documents product variants and merchant inventory exports, but access still requires the appropriate permissions.[4][5] Neither API availability nor a product mockup proves demand or profitability.
Key Lessons
- Keep the original job specific. Favorite brands and relevant sizes define a clearer experiment than an unconstrained AI shopping assistant.
- Separate inventory truth from recommendation quality. An attractive offer is unusable when its exact variant is unavailable or its source is stale.
- Do not invent the failure. Missing economics and outcome details require an open verdict, not a confident shutdown story.
- Compare the current alternative. Google already offers size-aware price tracking. Merchant workflow and authorized data must earn a rebuild’s place.