# Free Lunch for User Experience: Crowdsourcing Agents for Scalable User Studies

arXiv: https://arxiv.org/abs/2505.22981 · HTML: https://mlsanity.com/paper/2505.22981/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use.
- Builds on 33 of its 75 references that are in the map.
- First posted on arXiv 29 May 2025.

## Placed in

- LLM agents and tool use / [Social simulation and multi-agent debate](https://mlsanity.com/agents/social-simulation-and-multi-agent-debate/)

## Builds on (most cited first, 30 of 33)

- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://mlsanity.com/paper/2308.08155/) (2023)
- [Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference](https://mlsanity.com/paper/2403.04132/) (2024)
- [Personalizing Dialogue Agents: I have a dog, do you have pets too?](https://mlsanity.com/paper/1801.07243/) (2018)
- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studies](https://mlsanity.com/paper/2208.10264/) (2022)
- [Scaling Synthetic Data Creation with 1,000,000,000 Personas](https://mlsanity.com/paper/2406.20094/) (2024)
- [Using cognitive psychology to understand GPT-3](https://mlsanity.com/paper/2206.14576/) (2022)
- [LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals](https://mlsanity.com/paper/2411.10109/) (2024)
- [From Persona to Personalization: A Survey on Role-Playing Language Agents](https://mlsanity.com/paper/2404.18231/) (2024)
- [Large Language Models for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineering](https://mlsanity.com/paper/2305.03403/) (2023)
- [Artificial Artificial Artificial Intelligence: Crowd Workers Widely Use Large Language Models for Text Production Tasks](https://mlsanity.com/paper/2306.07899/) (2023)
- [The Programmer's Assistant: Conversational Interaction with a Large Language Model for Software Development](https://mlsanity.com/paper/2302.07080/) (2023)
- [Roleplay-doh: Enabling Domain-Experts to Create LLM-simulated Patients via Eliciting and Adhering to Principles](https://mlsanity.com/paper/2407.00870/) (2024)
- [Generative Students: Using LLM-Simulated Student Profiles to Support Question Item Evaluation](https://mlsanity.com/paper/2405.11591/) (2024)
- [Mind the Value-Action Gap: Do LLMs Act in Alignment with Their Values?](https://mlsanity.com/paper/2501.15463/) (2025)
- [You don't need a personality test to know these models are unreliable: Assessing the Reliability of Large Language Models on Psychometric Instruments](https://mlsanity.com/paper/2311.09718/) (2023)
- [From Role-Play to Drama-Interaction: An LLM Solution](https://mlsanity.com/paper/2405.14231/) (2024)
- [UXAgent: An LLM Agent-Based Usability Testing Framework for Web Design](https://mlsanity.com/paper/2502.12561/) (2025)
- [Evaluating Cultural Adaptability of a Large Language Model via Simulation of Synthetic Personas](https://mlsanity.com/paper/2408.06929/) (2024)
- [LLMs generate structurally realistic social networks but overestimate political homophily](https://mlsanity.com/paper/2408.16629/) (2024)
- [LLM-as-a-tutor in EFL Writing Education: Focusing on Evaluation of Student-LLM Interaction](https://mlsanity.com/paper/2310.05191/) (2023)
- [AI for social science and social science of AI: A Survey](https://mlsanity.com/paper/2401.11839/) (2024)
- ['Simulacrum of Stories': Examining Large Language Models as Qualitative Research Participants](https://mlsanity.com/paper/2409.19430/) (2024)
- [TeachTune: Reviewing Pedagogical Agents Against Diverse Student Profiles with Simulated Students](https://mlsanity.com/paper/2410.04078/) (2024)
- [PATIENT-Psi: Using Large Language Models to Simulate Patients for Training Mental Health Professionals](https://mlsanity.com/paper/2405.19660/) (2024)
- [Towards Process-Oriented, Modular, and Versatile Question Generation that Meets Educational Needs](https://mlsanity.com/paper/2205.00355/) (2022)
- [Proxona: Supporting Creators' Sensemaking and Ideation with LLM-Powered Audience Personas](https://mlsanity.com/paper/2408.10937/) (2024)
- [Attributions toward Artificial Agents in a modified Moral Turing Test](https://mlsanity.com/paper/2406.11854/) (2024)
- [Simulated patient systems powered by large language model-based AI agents offer potential for transforming medical education](https://mlsanity.com/paper/2409.18924/) (2024)

## Built on by (most cited first, 1 of 1)

- [Not There Yet: Evaluating Vision Language Models in Simulating the Visual Perception of People with Low Vision](https://mlsanity.com/paper/2508.10972/) (2025)

## Abstract (arXiv)

User studies are central to user experience research, yet recruiting participant is expensive, slow, and limited in diversity. Recent work has explored using Large Language Models as simulated users, but doubts about fidelity have hindered practical adoption. We deepen this line of research by asking whether scale itself can enable useful simulation, even if not perfectly accurate. We introduce Crowdsourcing Simulated User Agents, a method that recruits generative agents from billion-scale profile assets to act as study participants. Unlike handcrafted simulations, agents are treated as recruitable, screenable, and engageable across UX research stages. To ground this method, we demonstrate a game prototyping study with hundreds of simulated players, comparing their insights against a 10-participant local user study and a 20-participant crowdsourcing study with humans. We find a clear scaling effect: as the number of simulated user agents increases, coverage of human findings rises smoothly and plateaus around 90%. 12.8 simulated agents are as useful as one locally recruited human, and 3.2 agents are as useful as one crowdsourced human. Results show that while individual agents are imperfect, aggregated simulations produce representative and actionable insights comparable to real users. Professional designers further rated these insights as balancing fidelity, cost, time efficiency, and usefulness. Finally, we release an agent crowdsourcing toolkit with a modular open-source pipeline and a curated pool of profiles synced from ongoing simulation research, to lower the barrier for researchers to adopt simulated participants. Together, this work contributes a validated method and reusable toolkit that expand the options for conducting scalable and practical UX studies.