Alice Chatbot

Designing Alice Chatbot to Optimize Front Office Operations

→ Role
Product Designer

→ Platform
Web

→ Year
2022 - 2023

About the product

ALICE is a comprehensive product that improves operational efficiency and communication within the hospitality industry, contributing to better productivity and higher guest satisfaction. The front office department benefits from ALICE's chat solution, using it to communicate directly with guests.

High Level Business and User Goals

  • 40% decrease in active guest SMS conversations, giving hotel staff more time to focus on complex tasks that demand their expertise.

  • 75% improvement in average conversation completion time, reduced from 4 minutes to just 1 minute.

  • Position the Alice Chatbot as a Competitive Advantage.

Team

  • Product Designer: Fernando Chaves

  • Product Manager: Kenya Puig

  • UX Writer: Catherine Montesdeoca

  • Engineers: Fernando Silva, Konstantin Guryev, and Doğan Çelik

My Role

As Product Designer, I led the design process from conception to execution, collaborating closely with cross-functional teams including, product managers, engineers, and business units. Together, we crafted and enhanced an innovative chatbot experience.

Problem Hypothesis

Front Office Agents spend significant time handling repetitive guest inquiries, such as check-in times and Wi-Fi passwords, limiting their ability to focus on complex, high-value tasks that require their expertise.

Outcome

By implementing a messaging solution to handle FAQs, we reduce active guest SMS conversations by 40%, improve conversation completion time by 75%, and position Alice Chatbot as a competitive advantage.

User Research

Validating the Hypothesis

To validate our hypothesis, we undertook 3 key initiatives:

  • Desk Research: Documented examples of repetitive guest questions.

  • Data Analysis: Ran a query to analyze the percentage of guest messages containing predefined keywords related to common questions, such as 'check-out' and 'check-in.'

  • Qualitative Research: Conducted 12 in-depth interviews with Concierges and Front Desk Agents.

Initial Findings

  • 26% of guest messages contains keywords related to common questions, such as "check-out" and "check-in."

  • It's common for Front Office desks to close late at night, highlighting the need for a feature that supports users during these hours.

As a result of the research, we gained greater confidence in the project's direction and in our ability to solve a real problem.

Concept

Next, I worked on the user flow, ensuring that all user interactions across different permission levels were covered, both on the backend and frontend.

Following that, I worked on the UI elements to create intuitive and seamless interfaces that align with the user flow, ensuring a consistent and engaging experience for users at every interaction point.

User Flow

Next, I worked on the user flow, ensuring that all user interactions across different permission levels were covered, both on the backend and frontend.

Following that, I worked on the UI elements to create intuitive and seamless interfaces that align with the user flow, ensuring a consistent and engaging experience for users at every interaction point.

Usability Test

I convinced the Product Manager to validate the design and experience through a usability test with active users, highlighting its a way to mitigate risks.

Using Maze for unmoderated testing, we gathered over 100 responses. Participants completed tasks and shared comments, with most navigating with ease. Insights emerged from three distinct user groups:

  • The largest group believed the feature would simplify their workload and fully embraced the chatbot's capabilities.

  • A 'skeptical but optimistic' group was open to trying the chatbot but expressed concerns about its ability to handle guest requests accurately.

  • The smallest group doubted the chatbot’s effectiveness and preferred direct guest communication for a more personalized experience.

Next Steps

  • The feature is currently in Beta, and we are actively collecting user feedback to identify and prioritize improvements before it becomes generally available (GA).

  • The chatbot will be enhanced with additional functionalities tailored to meet evolving user needs and address more complex scenarios.

Learnings

  • Large-scale projects like this demand teamwork to address complex challenges, such as developing Escalations Logic: when the chatbot can't respond, the card turns yellow, prompting user action.

  • Design critiques are essential for improving any design.

  • Continuous user validation ensured the solution met real-world needs.

  • Collaboration with developers streamlined implementation and resolved technical challenges.

⤷