YOOX NET-A-PORTER GROUP (YNAP)

is the global partner of choice of the world's leading luxury brands for all online distribution channels

  • 24 Million monthly unique visitors worldwide
  • Customers in over 180 countries
  • More than 2 million high spending customers
  • Combined net revenue of over £1.3 billion
Next

Why a need for a chatbot?

With the focus of providing the best online shopping experience,
it is vital to answer customers’ queries in a timely manner.


  • 6 million women shop and browse pages of the site each month
  • A highly engaged loyal audience with 85 million monthly page views
  • 8.7 million items stocked worldwide and over 7 million orders shipped
  • 3.4 million followers on social media

As YOOX NET-A-PORTER GROUP has such diverse and extensive business areas, our client and our team will focus only on NET-A-PORTER, which provides online retail services targetted at women.

Currently, NET-A-PORTER uses a third party tool that connect users to the customer support team through a live-person chat messenger.

However, customer support will only appear when a user triggers certain actions (looking for sizing details or struggling to complete the order) as it is operating by real people. Moreover, it will only appear when customer support is available, which means that there will be instances where not all customers have access to the chat messenger. The point is that a query not answered might result in the loss of a potential purchase/customer.

With our Customer Service Chatbot employing Natural Language Processing, Artificial Intelligence and Machine Learning, it will be able to answer most of the customers’ queries.

It will act as a filter between the customers and human fashion advisors, who are on-site NET-A-PORTER employees.

This will allow NET-A-PORTER to better control and customise the customers’ experience on the website and analyse available query data to develop new business insights.

Project Abstract

Answering customer’s queries in a timely and efficient manner is important in providing the best online shopping experience.

A chatbot will be the perfect application to tackle this problem. We built the customer service chatbot using IBM Watson Conversation and Rank and Retrieve service with NET-A-PORTER APIs.

It is intended to serve NET-A-PORTER customers and act as a filter between them and the company’s on-site fashion advisors.

Next

Product Recommendation

The chatbot is able to recommend products

  • Category
  • Brand
  • Size
  • Colour
  • Price

Customer Care

Get answers for frequently asked questions regarding delivery, returns, payments, etc.

Next

Collaboration

The chatbot service is accessed with a REST API that we had developed.

Microsoft HoloLens

The API can be integrated with the other two NET-A-PORTER teams whose Microsoft HoloLens projects are able to use Speech-To-Text to implement our service.

Natural Language

Users will be able to request for products and FAQ answers using natural language.

YOOX and MRPORTER

This chatbot service architecture can also be integrated into other websites within the NET-A-PORTER Group such as YOOX and MRPORTER.

We are 2nd year Computer Science undergraduates from University College London.

Wayne Tsui

wayne.tsui.15@ucl.ac.uk
Team leader, Project Manager,
Chief Editor, Back-end Developer

Aouss Sbai

aouss.sbai.15@ucl.ac.uk
Sub-editor
Developer

Jason In

jie.in.15@ucl.ac.uk
Front-end Developer
UI/UX Researcher