How can we enhance Pleo's expense categorization to make it more intuitive for users?

Introduction

Enhancing Pleo's expense categorization to make it more intuitive for users is a critical challenge that directly impacts user experience and operational efficiency. As we dive into this product improvement case, I'll outline a strategic approach to address this challenge, focusing on user needs, pain points, and innovative solutions.

Step 1

Clarifying Questions (5 mins)

• Looking at Pleo's position in the expense management market, I'm thinking about the scale and complexity of transactions we're dealing with. Could you give me an idea of the average number of transactions per user per month, and the diversity of expense categories we currently support?

Why it matters: This helps us understand the scope of the categorization challenge and whether we need to focus on breadth or depth of categories. Expected answer: Average of 50-100 transactions per user monthly, with support for 30-50 standard expense categories. Impact on approach: Higher transaction volume would push us towards more automated solutions, while a vast array of categories might require a more flexible, user-driven approach.

• Considering the evolving nature of work and expenses, especially with remote work trends, I'm curious about our user base composition. What's the split between traditional office-based companies and more distributed or remote-first organizations among our customers?

Why it matters: Different work models may have distinct expense patterns and categorization needs. Expected answer: 60% traditional office-based, 40% remote or hybrid. Impact on approach: A significant remote user base might lead us to focus on more flexible, location-agnostic categorization systems.

• Given the critical nature of financial data, I'm thinking about the balance between automation and user control. What level of manual oversight or correction are users currently exercising over automated categorizations?

Why it matters: This indicates user trust in our current system and areas where human intervention is still necessary. Expected answer: Users manually adjust about 20-30% of automated categorizations. Impact on approach: High manual adjustment rates would suggest a need for improved machine learning models or more user-friendly correction mechanisms.