Pan Macmillan Chatbot for Book Recommendation [Case Study]

Pan Macmillan chatbot is a personal agent that simplifies book choice

Pan Macmillan chatbot

Can a chatbot work as a book recommendation assistant? Pan Macmillan Publishing in search of an answer to this question, familiar to many publishers, proved it can.

Once the company found itself piled up with frequent customer requests on what book to get as a gift or just read, the publisher decided to bring value to its customers by providing them with a personal assistant to chat with and choose a book, and at the same time, automate customer communication management.

Pan Macmillan came up with a one-box solution that would provide millions of customers with a personal agent, and address repeated customer queries efficiently. This is how the Pan Macmillan chatbot for Facebook Messenger was born.

Chatbot is able to tailor publisher’s offers to customer needs and improve its recommendations.

Personalization

Not only does Pan Macmillan chatbot automate recommendation service within the messenger’s interface recognizable for customers. It also treats every customer personally and tailors its recommendations specifically to the customer’s needs.

Pan Macmillan bot is a chatbot solution built in cooperation between BAM Mobile and Digiteum technology agency. Chatbot talks to customers like a human and helps them choose and order books online via a conversational interface.

API integrated with two book recommendation and rating services – GoodReads and Supadu – and connected to Amazon, it provides customers with relevant information on the books they are interested in. As a smart personal agent, the chatbot simplifies book choice and comes up with recommendations on the spot answering even the simplest user requests like Horror or Biography.

Chatbot applies its recommendation algorithm to broadcast and keep customers updated.

Smart book recommendation assistant

Chatbot guides users along the way narrowing book choice to certain categories: author, genre, title or reader’s age, and recommends books based on input combinations like “adventure child 5-7 years old.”

Once the book is chosen, the bot offers users to redirect them to Amazon and make online order without additional efforts. This is how chatbot becomes an additional channel of communication between a brand and its consumers and drives sales, thus increasing revenue and profits.

Primarily, Pan Macmillan chatbot applies rules to analyse user’s input and respond with the best possible scenario. Relying on a rich storage of trigger keywords, the bot understands customer’s preferences, memorises them, and further uses this information to build its book recommendations intelligently.

Chatbot applies its recommendation algorithm to broadcast and keep customers updated.

Broadcasting

What happens when the chatbot comes across the input it doesn’t recognize? It does exactly what a real person would do – replies with a joke. And memorizes failed inputs to let the development team improve its responsivity and, thus, customer experience in future. In other words, not only does the bot take into consideration the user’s personality, but also has one of its own.

Another function of the Pan Macmillan bot is broadcasting. Chatbot applies its recommendation algorithm, customers’ preferences, and their previous inputs to be the first to tell customers about the publisher’s offers.

Therefore, the chatbot acts as a smart agent able to tailor the publisher’s offers to customer needs instantly and improve its recommendations along with the history of communication with each customer individually.

Pan Macmillan uses chatbot to answer repeated customer queries efficiently and give value to readers.

Highlights

Automated human-like conversation between the brand and its clients via a popular channel of communication with billion users (messengers).
User engagement statistics: conversation duration starts from 3 minutes and takes at least 15 messages.
Personalized book recommendations built on user input analysis and data from third-party services.
Integration with third-party recommendation and rating services to provide users with relevant content.
Connection and redirecting to the point of sale (Amazon).
Product broadcasting and availability updates are sent to customers directly via messengers.
User input and feedback recording for further failure analysis and customer journey improvement.