Availing accurate and extensive fashion tags is imperative for any fashion discovery platform. Extensive tagging of minute fashion details delights fashion-centric customers and they end up shopping a full bag of products. One of the leading fashion social networks approached us for the creation of product tags for 20 different attributes like color fill type, sleeve, neck, embellishments, construction, etc. Each attribute had 7 to 26 different possible applicable values.
SCALE: 1 million tags for 60,000 top selling products
DURATION: 14 days
- Playment understood definition for each of the allowed values across all the attributes.
- Since the Tasks were complicated, Playment came up with innovative Task types to breakdown the complex problems into a series of simple Tasks.
- The Tasks were targeted to the right kind of Players personas. For this case young women on the platform identified by Playment’s self-learning player selections logic (PSL).
- Depending upon the level of subjectivity a particular Task was routed to anywhere in between 3 to 10 players on the platform.
- Post-processing of result collected via crowdsourcing was done by in-house data experts.