XPend

Fintech · Behavioural AI · Product Design · Harvard · 2024

Behavioural DesignConversational AIMobile UXConsumer PsychologyPoint-of-spend
The problem

Most personal finance apps assume people overspend because they lack information. The research consistently shows otherwise. Impulsive spending is emotional - triggered by stress, boredom, social comparison - and happens in moments when no one is opening a budgeting app. Showing someone where their money went, after the fact, does not change the decision that was already made.

Emotional
the actual trigger, not information gaps
Point-of-spend
intervention, not retrospective
Conversational
primary interface, not dashboards
Harvard
product management studio
The insight
A finance app that shows you where your money went is a diary. One that changes how you spend it in the moment is a product.
My approach

The product decision that defined everything: no charts on the home screen. Every other finance app leads with a graph of past behaviour. XPend leads with a conversational input, because people engage with their money more honestly in dialogue than in front of data.

The design was built backwards from the emotional state users are actually in when they overspend - not the state they're in when they open a budgeting app with good intentions. That meant the interface had to be frictionless, non-judgmental, and capable of a single useful interaction in under 30 seconds. Every feature that required more than that was cut.