Movie Theater chatbot

Conversational chatbot to order movie tickets using your mobile phone.

Summary

For a movie theater, called Natlab, based in Eindhoven, The Netherlands, I researched, prototyped and tested a conversational chatbot. The goal of this project was to discover if and how an effective mobile chatbot could be designed that supports users (scoped to students) to find interesting movies and book tickets.

University project Design & User Research, Prototyping, User testing

2019

Research & Design process

Problem Space

Chatbots are useful to get automated results from large directories. Together with the Natlab Movie Theater, I explored a chatbot to find movies to watch. I started the project by meeting the client, conducting background research, and exploring audiences and platforms.

Natlab Movie Theater in Eindhoven, the Netherlands

User Research

To better understand the user, I conducted qualitative research through interviews and quantitative research by handing out questionnaires for one week to visitors who went to see a movie. The aim was to discover what users usally do to find an interesting movie to watch and how they do this specifically at Natlab.

"Reviews and recommendations from friends are my primary source for choosing a movie at Natlab."

"I would only go to the cinema if I have a friend that wants to join, and (s)he has to like the movie too of course."

"Choosing a movie really depends on my mood. Sometimes a movie can be too heavy for the moment."

User Journey

The journey for each user was very different. For example, whereas watching the trailer is a must for some, others are very strict and do not want to get any spoilers through trailers or descriptions. The results of the user research were mapped into a simplified user journey. This gave insights into the movie-finding process of the user and it identified important parts in which the chatbot could help.

Conversation flow

To facilitate all users, the chatbot had to be very flexible. I created a converation flow to visualize the logic, responses, and behaviour of the chatbot.

Prototyping

Then, I developed the chatbot according to the conversation flow. I used Flow XO for prototyping.

User Testing

I tested the prototypes with users by giving them two tasks to complete: (1) 'Think about a day this week you would like to go to the movie theater. Use the chatbot to complete your entire process from beginning till end', and (2) 'Think about a friend with whom you have been to the movies recently. You are planning on going again tomorrow after dinner. Use the chatbot to complete your entire process from beginning till end.' I observed them and did follow-up interviews afterwards.

Results

The findings of user testing are captured in a set of 9 guidelines (about topics such as language tone and conversation flow) to provide directions for further development, as well as some additional conclusions which I presented to the client.

75% of respondents would be interested in using such a chatbot (albeit after some major technical upgrades).
From a content point-of-view, the prototype already covered most parts of the user-journey succesfully. Respondents primarily valued the chatbot's speed compared to a regular website.