CX Optimised Chatbot

Year

2024

Client

Propella.ai

Role

UX/UI

Deliverable

How I Guided Developers to Enhance Customer Experience in an AI-Driven Chatbot

Visit Live Website

How was I successful

I was able to work alongside the tech and dev team, to develop a more useful, ai powered chatbot product that was more customer experienced focusedI.

I used content design to orchestrate the flow of the chatbot in a way that made sense to usersI championed for the chatbot to be utilised on the companies own site as a demoI successfully used storytelling to re-work communication that was utilised across multiple platforms

Challenges overcome

I wore many hats for this project - User Interface Designer, Content Designer, Content Strategist, Marketing Guru and User Experience Designer.

I shaped the chatbot conversation flow by carefully balancing discovery questions to build user trust, providing valuable responses to product inquiries, and seamlessly guiding users to the desired outcome.

My approach prioritised crafting an engaging and natural interaction that not only met users' needs but also aligned with key business objectives.

Key Learnings

Working closely with the AI development team gave me valuable insights into integrating AI with traditional technology to elevate products.

This experience also clarified my preferred working style—I enjoyed taking a product design perspective and exploring the project from various angles.

Understanding LLM-structured chatbots provided me with a solid foundation to communicate their benefits and discover new use cases.

one

Unrefined AI ChatBot Product

two

Lack of market research

three

Lack of user optimisation

Business Problem

Propella.ai was on the forefront of releasing a new service offering, targeting business owners who wanted to take the leap into automating customer service with AI.

The main challenge in developing AI solutions was effectively communicating their benefits to diverse audiences, many of whom were unfamiliar with AI or unsure how to integrate it into their business.

We aimed to create an AI-powered chatbot for a wide range of industries, from e-commerce to SaaS, which required market research to identify the unique challenges each industry faced when engaging with customers. The business goal was to clearly differentiate these AI chatbots from standard bots, showcasing their advanced capabilities.

With a successful product ready to be marketed, Propella could begin reaching out to prospective clients with a clear use case, demonstrating how the chatbot could enhance customer engagement, streamline processes across various indiustries

one

Understand Propella website client journey

two

Research client journeys in other industries

three

Create a go-to-market ready product

Measuring Success

My role in this project began by testing the initial chatbot version on the website to understand its role within the user journey. I then designed the UI to align with the brand’s identity, enhancing conversational flows through user testing for a more natural experience.

I also conducted market research across industries, creating mock customer journeys to showcase the chatbot's potential. A key challenge was effectively communicating its value, especially as chatbot capabilities are expected to become standard within the next five years.

‍

Product Roadmap

Communicating the value of customer experience

I collaborated with the developers after testing the initial demo version to discuss potential improvements. I recommended incorporating a combination of predictive text and typed responses to enhance interaction speed. While the chatbot delivered valuable detail and insights, it was crucial to maintain user engagement throughout the experience

DEFINE

No items found.

RESEARCH

Market research

The initial demo version was intended to be live on the Propella website, but to fully productise the chatbot, we needed to revisit the concept and explore additional industry use cases for broader promotion. I conducted research into existing chatbots, identified their limitations, and used these findings to highlight the unique solutions our product could offer. At the request of the CIO, I began with a use case for real estate agents, leveraging our established connections in that industry.

No items found.

MVP

Developing an MVP for Four Key Use Cases

The decision was made to start with a simplified version of the chatbot while the underlying technology was still being developed, rather than pursuing a more complex system that would require a full-scale application, particularly for the real estate sector.

We ultimately selected four distinct use cases: ECommerce; Technical support for SaaS; Product Recommendations and Lead Generation

No items found.

USER TESTING

Validating the logic

I played a key role in developing a retail product recommendation chatbot designed to build trust and increase customer conversions through authentic and helpful interactions.

I recommended adding two layers of product knowledge questions to boost credibility before making recommendations. This refined flow became an effective demo for prospective clients, highlighting how AI can drive meaningful customer engagement

No items found.

COMMS

Building the story

I leveraged my deeper understanding of how the chatbot works to create a marketing deck for investors and stakeholders. I emphasised the importance of a clean, modern UI—similar to ChatGPT's style—to attract and engage customers effectively

No items found.

MOCK-UP

Final Design

In the final UI designs of the chatbot, I transformed the initial, basic interface into a modern and visually engaging experience. The original design was functional but lacked the visual appeal and intuitive elements expected by today’s users.

I focused on updating the overall look and feel, incorporating a clean, contemporary design aesthetic that emphasised accessibility and user-friendliness.

No items found.

Next Steps

1. Monitor Performance Metrics and Gather User Feedback - Collect user feedback to understand pain points or areas for improvement. This helps identify any gaps in user experience and refine the chatbot’s performance accordingly

2. Expand Use Cases and Integrate with Other Systems

‍

Final Notes

I gained a lot from working on this product. It was lightweight with a quick turnaround, and the agile team made implementing changes seamless. This experience helped me build confidence in trusting my own judgment to shape the product, and it allowed me to step into more of a leadership role, similar to that of a Product Manager

During the project, I learned the importance of user-centric design, ensuring the interaction was authentic, helpful, and trustworthy, particularly when dealing with an unfamiliar technology like AI.

‍

Next Project

View next project