Context

Currently ,of the users who are landing in the Mix workflow a very small number is actually initiating image cutouts and are able to get a satisfactory result to get an export. While engineering looked to fix the object identification model, design focused on making the experience more seamless.

Current pain-points

Data based insights

• The data tells us that from 100% of the user coming to the mix workflow, 80% of the user complete the first stage of onboarding and 60% of the users tap on cutouts.
• Of the 60% of users selecting on cutout, 46% of them finally start the onboarding
• There is a 15% drop of users, from the set of users who finally start onboarding.
• Of the cumulative number who finally land on we have the last slide elaborating on what % of users tap on which pill.
• 45% of the users tap on add custom cutouts, but the throughput From custom cutout stands at 15%
• An average of 23% of users cumulatively add/delete layers when in cutout with propensity of paid users ( 30%) doing more than free

Competitive Audit

Most of the applications audited follow an object selection model, with key actions to add/remove elements in the model in the L1 level.

Only one application only gives the scope to the user to select objects manually.

Inplace cutout: the toggle that allows the user to view the selection and the selection wrt to the composition is a key component in all the competition

Photoleap & photo leaf are the two applications which has a layer workflow