Marine Layer — Alex Rosales

Marine Bot Features

Marine Bot greeted the user with the options for what she could do.

Feature Cards

Marine Bot would serve up cards for each feature, clearly laying out next steps.

Shipment Tracking

With the shipment tracking card, users would not need to click out of the Marine Layer experience to know when their product would arrive.

Return Card

Users would not need to select previously purchased product within another experience. The return card would serve up the most recently purchased product on a user's account and provide a simple experience to return the item.

MARINE LAYER - CUSTOMER SERVICE CHATBOT PROTOTYPE

Marine Layer is an apparel retail startup committed to bringing the softest clothes to the world. I partnered with the company on a capstone project for a product bootcamp I attended in San Francisco. I created a high fidelity prototype in Sketch and InVision to provide customer service via chatbot.

What problem was I trying to solve?

For customers:

Long wait times/no response to email for customers with a customer service inquiry
Frustration in tracking and return experience
No status of sold out items

For Marine Layer:

Understaffed customer support
- Marine Layer customer support is only available through email and behind that email is a single employee whose job is part customer service and part operations.

What did I design?

I created a high fidelity prototype in Sketch and InVision to provide customer service via chatbot in Facebook Messenger or other messaging platform. For my capstone, we were asked to make a MVP prototype with explanation of future features based on user stories.

For the MVP, I included the following features:

Track and order
Return an item
Request sold out item
Speak to a human

Future features included:

Messages to users when an order has shipped/been delivered
Messages to users when sold out item is in store/online

Discoverability was apparent as users are provided with a clickable list of all features. The idea was to show users what the chatbot could do and be able to do these functions quickly and easily.

Should a user type a command not listed, they would be prompted to connect with a human. Their query would be routed to customer service staff, who could reply within messenger.