Work Verizon

What if your research participants aren’t real?

Teams decide where to focus without research, because CMI cannot get there before priorities do and a Figma persona cannot be tested against. So I built synthetic users: sixteen published prompt frameworks benchmarked against 169 people.

23
Use cases the business asked for
16
Instruction frameworks tested
3
Synthetic users live
Role
Lead Experience Designer II
Team
Experience Architecture, Identity Architecture
Company
Verizon
Period
2026 — Present

AI in this work

  • Decks built with AI from the meeting notes, so each session opened current on the last.
  • Current-state and future-state storyboards generated from the customer story, then redrawn for the room.
  • Initial card sorts and tree tests run against the synthetic users themselves, to scope what Consumer & Market Insights would then be asked to study.

Protected

Confidential information

Unlock this case study to see the results of this work by contacting me for the password.

What's inside

Sixteen published prompt frameworks, each given the same persona and the same two tasks, ranked against what 169 real people did. Nine tools built on the winner, two of them killed, and what a synthetic user turned out to be bad at.

Request the password

Usually within a day Or write to nachapman@gmail.com

Top