TheySaid is an AI user-testing platform. It gets feedback on your designs, prototypes, and live products from synthetic testers trained on your customer data, from human testers, or from both. You run user tests, AI-moderated interviews, surveys, forms, and polls in one builder. AI then summarizes themes, recommends fixes, and creates action items. It suits UX researchers, designers, and product managers who need customer validation quickly without scheduling sessions.

• Synthetic audiences: Test with AI personas built from your own customer data instead of generic personas. TheySaid says its base model draws on 100,000+ hours of usability sessions.
• Unified test builder: Build user tests, interviews, surveys, and polls in one interface, so you don't need a separate tool for each research method.
• AI test moderator: AI runs your sessions, asks follow-up questions, and adjusts questions in real time. You don't have to schedule or moderate them yourself.
• AI project creator and Teach AI: Describe what you want to learn and AI builds the whole project. You can also train it on your goals, users, and research needs.
• AI analytics and Ask AI: AI summarizes learnings, themes, and recommendations automatically. You can ask questions across one project or many to compare insights.
• Voice conversations and 70+ languages: Participants can answer by speaking while AI reads the questions aloud. Support for more than 70 languages lets you reach users worldwide.
• Flexible distribution: Share a project by link, email, in-app popup, website embed, or QR code, or recruit through the panel to reach the right audience.
• AI builder and ticketing integrations: The TheySaid MCP brings feedback and fix prompts into tools like Claude Code, Cursor, Replit, v0, Bolt, and Lovable. Ticket Testing starts customer validation when a Jira ticket finishes design.

• Fast turnaround: Reviewers point to how quickly they can run AI-powered user tests and get synthesized insights, compared with recruiting and moderating sessions by hand.
• One platform for many methods: You can run usability tests, interviews, surveys, forms, and polls in one place instead of paying for several separate tools.
• Free entry point: The Free plan needs no credit card and includes 5 AI user tests and 50 AI surveys each month, so you can try the workflow before you buy.
• Mix synthetic and human feedback: You can test early ideas with synthetic testers, then confirm them with real people. TheySaid's own studies report that the two groups caught a similar share of issues.
• Fits AI build workflows: Integrations with AI coding and design tools, plus Jira, Linear, ClickUp, and monday.com, put feedback into the tools your team already uses.
• Insights you can act on: Automatic recommendations, action items, and highlight-reel clips turn raw responses into findings you can share with your team.

• Only two public tiers: The pricing page lists a Free plan and a custom-priced Enterprise plan with nothing in between. Small teams that outgrow the free limits have to go through a sales conversation.
• Tight Free plan limits: The Free plan allows 1 workspace and 1 active test and uses your own users. That's enough to evaluate the product but not to run ongoing research.
• Key capabilities are Enterprise-only: Panel recruiting from the 16M+ participant pool, the custom API and webhooks, SSO/SAML, and CRM integrations (HubSpot, Zendesk, Workday) all require Enterprise.
• Extra costs for niche audiences: Recruiting hard-to-find testers, such as business owners or doctors, costs extra, so total spend can vary with how specialized your audience is.
• Synthetic testing is not a full replacement: Synthetic testers are AI simulations. Studies that need real behavior or regulatory evidence still need human participants, which raises cost and time.

• Validating prototypes before development: A product designer runs a usability test on a new checkout prototype with synthetic testers. They find navigation problems in hours and fix them before any engineering work starts.
• Checking AI-built apps: A founder building with Lovable or Cursor uses the TheySaid MCP to get feedback and fix prompts inside their AI tool. They can catch broken flows before real users hit them.
• Running interviews without scheduling: A UX researcher sets up AI-moderated interviews in several languages. Participants answer by voice whenever it suits them, and AI summarizes the themes across every session.
• In-app pulse checks: A product manager triggers an in-app popup poll after users try a new feature. The quick rating plus a follow-up question shows how the feature is landing.
• Validating Jira tickets: An enterprise team turns on Ticket Testing, so customer validation starts automatically when a Jira ticket finishes design. Testing becomes part of the delivery process.
• Cross-project research analysis: A research lead uses Ask AI to compare findings across several studies. Recurring usability problems surface without anyone reviewing transcripts by hand.