Synthetic Users
UX Audit
Design System
Capstone
CASE STUDY EN
★ Featured · UX Research with AI

Synthetic Users.
Research that runs in 30 minutes.

Define a persona, hand it to Claude, and run real research before you run real research — and learn when to trust it, and when it's lying to you.

Duration7 Weeks · 55 Hours Format9 On-demand Sessions OutputBehance Case Study ProjectAlmosafer UX Research
Concept
Method
Practical
Recap
02 / 14
The Tension

The Promise — and the Risk.

Pre-test every design decision against 5 personas in 30 minutes. No scheduling. No recruiting. No no-shows. The leverage is real — but so is the trap.

★ The Promise Accelerator

A research accelerator

Sharpen your thinking, catch blind spots, pressure-test a flow — all before you book a single real interview.

⚠ The Risk Hidden Trap

Plausible-sounding fiction

Teams who skip real research because the AI "already told them" what users think. Three weeks of fake confidence, one hour to be humbled.

Concept
Method
Practical
Recap
03 / 14
Concept

What is a Synthetic User?

An AI persona you define in rich detail, then interview or usability-test. Claude plays the character. You play the researcher.

★ 01 Roleplay

Directed roleplay

Use Claude's reasoning about a character to generate hypotheses and pressure-test assumptions — with a clear research purpose. Not creative writing.

★ 02 Agent Loop

Familiar pattern

A persona becomes a system prompt; you run a scripted conversation. The "agent" plays a user — not a software builder.

★ 03 Repeatable

30 min vs 3 weeks

Run the same interview with 5 personas. Iterate the prompt. Re-run. The leverage compounds with every cycle.

Concept
Method
Practical
Recap
04 / 14
Method

Two Modes. Same Persona.

Same Claude Desktop. Same persona. Different research moments — one for hypotheses, one for friction.

A
💬 Conversation Mode
Synthetic Interview

Open-ended questions about motivations, expectations, fears. Run before the persona sees the product. Surfaces mental models, hesitations, trust signals.

B
🌐 Live-site Mode
Synthetic Usability Test

Show a screen or real URL. Narrate a task, observe think-aloud reactions. Run inside the product via Playwright. Finds friction, labels confusion, exposes edge cases.

Concept
Method
Practical
Recap
05 / 14
Quality Driver

Quality comes from Persona Richness

Thin personas give you thin opinions. Rich personas give you specific, surprising, actionable reactions.

✕ Thin Bad Brief

thin-persona.md

"A 30-year-old woman in Riyadh. Shops online. Middle income."

Generic, safe, averaged opinions. Nothing sharp. Nothing you didn't already guess.

✓ Rich Good Brief

sara.md

"Sara, 34, Riyadh. Marketing manager, mother of two. Got burned last year by a fake Instagram seller — since then she only buys from big names."

Specific reactions: "the missing review count worries me", "I wouldn't trust a site that asks for payment before showing the return policy."

Concept
Method
Practical
Recap
06 / 14
Is This Real?

Yes — with receipts.

This isn't a ChatGPT gimmick. Industry research leaders are actively publishing about synthetic users — both sides of the debate.

★ Verified NN/g

The Balanced Guide

"Synthetic Users: If, When, and How to Use AI-Generated Research." Not dismissing, not endorsing. NN/g publishing on this signals the field is taking it seriously.

★ Strategic Bain

Strategic POV

"How Synthetic Customers Bring Companies Closer to the Real Ones." Treats synthetic users as legitimate augmentation of customer research.

Gartner-cited syntheticusers.com

85–92% parity

Gartner-cited tool. Claims 85–92% parity with real interviews on thematic overlap and depth of insight. Parity numbers exist — that alone makes it measurable.

⚠ Debate ACM Interactions

Both sides published

"The Synthetic Persona Fallacy" + "The Challenges of Synthetic Users." The research community is actively wrestling with this — live debate = real topic.

Concept
Method
Practical
Recap
07 / 14
⚠ Warning

The Documented Failure Case.

A healthcare design team's three-week synthetic research practice — flattened in a single hour with real humans.

3 weeks Synthetic-only

Three weeks of confidence

3wspent running ChatGPT-generated personas — building hypotheses, framing decisions, briefing stakeholders.
1 hour Real users

One hour with real humans

1hwith rural patients & physicians revealed complexity "far more than our artificial users could convey."
"UX research is fundamentally about hearing real human experience. Synthetic users cannot capture the why behind choices, or the contextual factors that actually influence behavior." — ACM Interactions, "The Synthetic Persona Fallacy"
The rule → Synthetic users are an accelerator, not a replacement. Pre-validation, hypotheses, sharper questions — yes. Statistics, accessibility, cultural nuance, high-stakes decisions — no.
Concept
Method
Practical
Recap
08 / 14
Template

The Rich Persona Template.

Copy this structure for any persona. The fields at the bottom — recent experience, emotional baseline, preferences — are where the signal lives.

## Sara Al-Otaibi, 34, Al-Malga district, Riyadh

Demographics: Marketing manager at a mid-size agency, 6 years in.
Mother of Layan (7) and Faisal (4). Bilingual — code-switches Arabic/English.

Context: Hybrid 9-to-5. Online shopping = mostly Namshi for kids' clothes
and Carrefour Online for monthly groceries. Has never tried panda.sa.

## Where the signal lives ↓

Recent experience: Last Ramadan, burned by a fake perfume seller on
Instagram (SAR 400 via bank transfer, vanished). Since then: no bank
transfers, no "new sites", checks for visible consumer-protection cert.

Emotional baseline: Tired. Doing this between two meetings, on her phone,
with her 4-year-old pulling her sleeve. Patience for forms = zero.
Reads 2- and 3-star reviews first (5-stars lie, 1-stars are drama).

Preferences: Saved addresses non-negotiable. Apple Pay, then Mada.
Abandons cart immediately if surprise delivery fee at final step,
or if forced to create an account before seeing the total.
Concept
Method
Practical
Recap
09 / 14
Target Site

The shared target: panda.sa/en

Online grocery store of Panda — one of Saudi Arabia's largest supermarket chains. Two missions, one real site, both before you ever talk to a real shopper.

Mission 01 Pre-visit

The Interview · before site contact

  • Surface expectations, fears, mental models
  • Trust signals, past experiences, what makes her leave
  • Output: 3 sharpened insights + 1 flat one
Mission 02 Live test

The Usability Test · inside the product

  • Walk full flow: search → cart → checkout
  • Capture friction, confusion, abandonment moments
  • Output: 5 UX gaps, ranked by persona-impact
Pre-warning → Panda shows a login popup on landing. Dismiss with Esc — clicking the X doesn't work (a UX bug itself). Full shopping flow is testable as a guest.
Concept
Method
Practical
Recap
10 / 14
The Cast

Meet the Personas.

Three reference personas to reuse — or invent your own. Three different mental models, three different sets of friction.

All travellers Cautious Price-hunter Custom ★ Recommended
S

Sara, 34

Cautious mom · Riyadh
  • Marketing manager, 2 kids
  • Shops online rarely
  • Burned by fake Instagram seller
  • Low patience for forms
  • Prefers Arabic UI
K

Khalid, 22

Price-hunter · Student
  • University student
  • Heavy Hungerstation/Jahez user
  • Expects apps to "just work"
  • Zero patience for friction
  • Compares prices every order
A

Amina

Your custom persona
  • Pick a different mental model
  • New emotional baseline
  • Different past experience
  • Fresh trust signals to test
  • Or — write your own

Three different mental models — three different sets of friction. That's the whole point of running multiple personas.

Concept
Method
Practical
Recap
11 / 14
Prompt Pattern

The Usability Test Prompt.

Copy this. Swap the persona. Swap the URL. Run it. Same template works for any product, any task.

You are [PERSONA NAME]. Stay in character throughout.
[PASTE RICH PERSONA HERE]

I want you to do a think-aloud usability test on this site:
URL: https://panda.sa/en

Task: Find essential groceries for a family of four for the week
(milk, bread, rice, vegetables), add them to cart, review totals,
and attempt to proceed to checkout.

Visit the site using Playwright. At EVERY screen:
  1. Describe what you see (in [PERSONA]'s voice)
  2. Say what you'd click next and why
  3. Call out anything confusing, frustrating, or surprising
  4. Note anything that would make you abandon the task

Take a screenshot at each key step.

# IMPORTANT
- Dismiss the initial login popup with the Escape key
- Stop at the final checkout login wall — do not create an account
- Stay in character. Don't summarize "what a user might think" — BE [PERSONA].
Concept
Method
Practical
Recap
12 / 14
Workshop

Run it yourself.

Two activities. Group work. Same Claude Desktop. Different research moments — interview first, usability test second.

Activity 01 ~20 min

Synthetic Interview

  • Pick one persona (Sara · Khalid · Amina)
  • Paste persona, ask Claude to stay in character
  • Interview about pre-visit expectations for panda.sa/en
  • Document 3 sharpened insights + 1 flat one
  • Share-out: one sharp, one flat, with the class
Activity 02 ~20 min

Synthetic Usability Test

  • Pick 2 personas, run the prompt twice
  • Walk full flow: first visit → search → cart → login wall
  • Per persona: 3 UX gaps + 1 surprise insight
  • Compare: where did they agree? Where diverge?
  • Agreement = stronger signal
Concept
Method
Practical
Recap
13 / 14
Decision Frame

When to use — when not to.

One app, two modes. Claude Desktop handles both. But knowing what NOT to use it for is the difference between a sharp researcher and a fooled one.

✓ Good for Use it
  • Pre-validation before committing
  • Hypothesis generation
  • Finding edge cases you missed
  • Pressure-testing assumptions fast
  • Early stakeholder pitches
  • Sharpening real interview scripts
✕ Never for Avoid it
  • Replacing real user research
  • Statistical claims
  • Accessibility / assistive-tech
  • Cultural nuance & lived experience
  • Anything high-stakes (medical, legal, financial)
  • Sign-off on a final design
The meta-lesson — cross-persona agreement is your strongest signal. Single-persona insights are hypotheses, not conclusions. Never stop there. — The whole point of running multiple personas
Concept
Method
Practical
Recap
14 / 14
★ Recap · End of session

Session 6 · Key Takeaways.

  • Synthetic users = accelerator, not replacement — accepted in industry, used honestly.
  • Rich personas beat thin demographics — recent experience, emotional baseline, and preferences are where the signal lives.
  • Two modes, one tool — synthetic interviews (conversation) and usability tests (Playwright + live site), both in Claude Desktop.
  • Cross-persona agreement = high-confidence signal — single-persona insights are hypotheses, not conclusions.
  • Always leave room for real humans — the "3 weeks synthetic, 1 hour real" failure case is why this rule exists.
📘 NN/g · Synthetic Users Guide 📊 Bain · Synthetic Customers 📝 ACM · The Synthetic Persona Fallacy 🛠 syntheticusers.com