About
A messy human, simulated. And audited.
Why it exists
Why Almost Human
Real people are made of tissue, water and biases. They contradict themselves, get interrupted, and say one thing while doing another. That mess is where most research insight lives. Synthetic personas are getting popular, so I wanted to see how close one could get to a messy human, and where it still falls short.
Method
How it works
Build a panel from five layers: demographics, behaviours, attitudes, biases, and the context someone is in when they meet your idea. Run any concept past them as interviews, a group discussion, or both. Then read the audit, which flags what a synthetic panel can't give you.
Limits
What it can't do
A synthetic panel can tell you why something should land. It can't show you whether it does. Its biases are set, not discovered, and its reactions are described, not observed. Use it to sharpen questions and spot gaps before fieldwork, not to replace talking to real people.
Who made it
Built by Radhika Venkatarayan, a UX researcher with over 25 years in qualitative research.
See my research portfolio