Theysaid

Users click a test link, share their screen and talk through tasks while an AI moderator watches and asks relevant follow-ups in real time. The session captures screen activity, voice feedback and behavioral timestamps. Once tasks are done, TheySaid generates task success rates, repeated-issue patterns and highlight clips without anyone editing footage.

Instead of a flat form, TheySaid asks rating scale, multiple choice and ranking questions and then follows up on the answers. A poll captures the number and the reason behind it in the same pass. The AI summarizes responses as they come in, so a team can read a clear conclusion rather than a wall of open text.

The AI leads interviews and adapts its questions to what the participant says, which is useful for win-loss research or customer discovery. Forms work the same way, so guided questions replace a rigid template. With two-way voice, users answer out loud and the AI reads questions back, which suits participants who would rather talk than type.

You describe the goal and the AI project creator builds the study, including question types and flow. Teach AI lets you upload context about your product, users and research goals so later questions stay on topic. This cuts the setup time on repeat studies and keeps a whole team working from the same briefing.

The AI analytics layer summarizes themes, learnings and recommendations on its own, then turns findings into action items you can push into connected tools. Ask AI answers questions across one project or several, so a stakeholder can ask what came up most often this quarter and get an answer tied to the actual responses.

Projects reach users through a shareable link, in-app pop-up, website embed, email or QR code. TheySaid supports more than 70 languages, so global products can run the same study across markets. Templates cover the common feedback types, and conditional logic keeps branching simple while AI follow-ups stay in play.