Start from a clearly defined user problem or need.
Every feature — including AI capabilities — must trace back to solving a real, validated issue discovered during research.
This prevents “AI for AI’s sake” and anchors all work in meaningful outcomes.
A problem-first mindset ensures the team builds what users actually need, not what technology simply allows.
Framing ideation around “How Might We” questions maintains focus on solving the right challenge.
Decisions should always be supported by data, insights, or experiment results.
From personas to feature prioritization, teams rely on user evidence over opinion.
AI tools provide faster analytics and richer data, but human interpretation remains essential.
Every hypothesis should be tested, measured, and refined — a design process inspired by the scientific method:
observe, hypothesize, test, and iterate.
Performance metrics like the North Star Metric (NSM) act as feedback loops that continuously inform design evolution.
Design for delight, not mere functionality.
The goal is to create experiences users genuinely love — intuitive, empowering, and emotionally engaging.
In AI-powered products, this means ensuring the technology feels human, helpful, and inspiring, not cold or robotic.
Small touches — a natural tone of voice, an elegant micro-interaction, or a satisfying animation — contribute to user joy and trust.