AI-moderated human testing
Participants follow a test link, complete guided tasks, and share their screen and thoughts. AI asks follow-up questions and records screen activity, voice feedback, and behavioral timestamps.
AI user testing with real customers and synthetic audiences for product teams.
TheySaid is an AI user research platform for product managers, designers, and researchers. It tests designs and working products with real participants or synthetic users informed by customer data, helping teams assess usability before shipping. It also supports surveys, interviews, and polls.
Teams can request validation from tools such as Claude Code, Cursor, v0, Replit, and Lovable. Tests capture how participants approach tasks, while AI analyzes feedback and returns findings that teams can use to revise their work.
Participants follow a test link, complete guided tasks, and share their screen and thoughts. AI asks follow-up questions and records screen activity, voice feedback, and behavioral timestamps.
Tune synthetic testers using selected call recordings, surveys, support tickets, or CRM data. Prebuilt market audiences offer a starting point, with Fast and Full modes for different session depths.
Generate summaries, task completion metrics, time-on-task statistics, recurring issue patterns, and prioritized recommendations. Timestamped transcripts and highlight clips help teams inspect the evidence.
Run tests through MCP or trigger validation when a ticket reaches a stage such as “Design Done.” Findings attach to the ticket, with suggested prompts for addressing identified issues in an editor.
Review an experience against its specification and best practices before participant testing. Teams can also check a working build against the spec before a pull request merges.
Collect feedback beyond usability sessions through additional research formats. Enterprise options include conditional logic, project templates, cross-project insights, and custom APIs and webhooks.
Use synthetic audiences to explore alternatives before committing to a direction. Compare buyer, end-user, and executive reactions to the same design in one report.
Have participants work through pricing pages, checkout steps, or onboarding flows to identify confusing information and points where they struggle to complete a task.
Test a proposed feature or prototype with existing customers or screened participants before investing further in development.
Use Critical Experiences to score stages of the customer journey after launch, and Experience Hub to bring feedback evidence into backlog tickets.
The user testing Free plan lists five AI user tests and 50 AI surveys per month, one workspace, and one active test. AI project creation, moderation, and analytics are included. No credit card is required. Enterprise pricing is custom.
For human tests, participants click a test link and complete guided tasks while sharing their screen and thoughts. The website states that downloads or technical setup are not required.
TheySaid positions synthetic testers for narrowing options and informing early design decisions, with results in minutes. Human testing uses customers or recruited participants to provide direct evidence before larger commitments.
Teams choose which customer data sources tune their synthetic audiences, how AI interprets that data, and what it can act on. Human tests also support screeners and a Response Quality Monitor to filter unsuitable participants and low-effort answers.
トラフィックデータは参考情報としてご利用ください。