TheySaid 4.0

Product Introduction

1. Definition: TheySaid 4.0 is an AI-powered user testing and customer validation platform that integrates synthetic user testing with traditional human panel testing. It functions as a continuous validation infrastructure within existing developer and product workflows.
2. Core Value Proposition: It exists to eliminate "Validation Debt"—the cumulative cost of shipping product decisions without customer feedback—by providing instant, actionable user evidence directly inside tools like Claude, Cursor, Jira, and design canvases, enabling data-driven decisions in minutes instead of weeks.

Main Features

1. AI-Powered Synthetic Testers: These are not generic AI personas. The technology is built on a base model trained on over 100,000 hours of real usability session data to replicate human hesitation and behavior. Synthetic testers can be fine-tuned using a company's own customer data (e.g., call recordings, support tickets, CRM data) and perform tasks that include screen recording and verbal think-aloud commentary, generating video-based evidence similar to a human session.
2. Integrated Human Testing Panel: The platform provides access to a managed panel of real human testers, which can be screened for specific demographics. The system handles recruitment, scheduling, and payment. Tests can be run with the same audience definitions and test builder used for synthetic tests, allowing for direct comparison.
3. Workflow Automation & MCP Integration: TheySaid operates as a background process within existing tools via MCP (Model Context Protocol) and native integrations. It can be triggered automatically by workflow events (e.g., a Jira ticket moving to "Design Done"), running validation tests and attaching the results directly to the ticket without manual intervention from the team.
4. Expert AI Reviewer: This feature provides an automated expert critique of a design or product build against its original specification and UX best practices. It acts as a quality gate to ensure alignment with intent before human or synthetic user testing begins.
5. Unified Test Builder & Multi-Perspective Audiences: A single interface allows teams to create tests that can be run across different "Audiences"—such as synthetic buyers, human end-users, and executive personas—on the same design. Results are consolidated into one report for comparative analysis.

Problems Solved

1. Pain Point: The slow, costly, and manual process of traditional user research creates a bottleneck, leading teams to ship features based on internal guesses rather than customer evidence, accruing "Validation Debt."
2. Target Audience: Product Managers, UX/UI Designers, Product Designers, Frontend Developers working in SaaS and digital product teams, especially those using AI-augmented development tools (Cursor, Replit) and design tools (Claude Design, v0).
3. Use Cases: Rapidly testing design prototypes from Figma or Claude Design; validating feature specs before development begins; conducting pre-release usability checks on live builds; continuously monitoring critical user journeys post-launch; A/B testing design variations with targeted user segments.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros: Dramatically reduces time-to-feedback from weeks to minutes/hours; significantly lowers cost per feedback session (from ~$50-100 per human tester to ~$1 per synthetic test); seamlessly integrates into existing dev/design workflows without requiring team-wide adoption of a new tool; enables high-frequency, iterative testing that was previously cost-prohibitive.
   • Cons: Synthetic testers, while highly accurate, are not a 100% substitute for human intuition and nuanced emotional feedback (they catch ~89% of issues vs. ~90% for humans). The platform's effectiveness for synthetic testing is partially dependent on the quality and quantity of a company's available customer data for tuning.

2. Key Alternatives & Differentiation:

   • UserTesting.com / UserZoom (Traditional Panels): TheySaid differentiates by offering integrated synthetic testing for rapid iteration and managed human panels, all within a workflow-native platform. Alternatives typically require separate tools, manual scheduling, and days/weeks for results, at a much higher cost per test.
   • Maze / Useberry (Unmoderated Testing): While these offer faster testing than traditional panels, TheySaid 4.0 differentiates with its deep AI-native synthetic testers that provide video/audio recordings (not just survey clicks and heatmaps) and its direct, automated integration into AI coding and design environments.
   • Building In-House Research: TheySaid provides a managed, scalable infrastructure versus the high operational overhead of recruiting, scheduling, and analyzing tests manually. It offers a 25x increase in validation volume for the same budget.

Frequently Asked Questions (FAQ)

1. How accurate are TheySaid's synthetic testers compared to real human testers? According to TheySaid's published side-by-side studies, synthetic testers identify 89% of total usability issues, compared to 90% identified by human testers. Notably, 10% of issues are caught only by synthetic testers, making them a valuable component for comprehensive testing.
2. What data is required to train a synthetic tester on my specific customers? You can connect data sources like customer call recordings, support ticket transcripts, survey responses, or CRM attributes. If no specific customer data is available, you can start with pre-built Audiences tailored to your market (e.g., "SaaS buyers") and refine them as you collect more data.
3. Does using TheySaid's AI mean my customer data is used to train public models? No. TheySaid emphasizes data governance, being SOC 2 and GDPR compliant. You control what customer data is used to tune your synthetic audiences, and the system is designed to keep this data within your defined parameters.
4. Can I test products in languages other than English? Yes, TheySaid 4.0 supports over 70 languages for tests, interviews, and surveys. It also features auto-translation, allowing you to run the same study across different geographic markets without manually recreating it.
5. How does the pricing for TheySaid's synthetic testing work? Synthetic tests start at approximately $1 per test, offering a radically lower cost compared to traditional human panel testing (often $50-$100 per user). This pricing model enables much larger sample sizes and frequent re-testing after design changes.