Questions people ask
before signing up.

HeySpeak is async voice feedback. You create a Magic Link with one question, send it by email or LinkedIn, and the recipient records a 60-second voice note in any browser. AI transcribes the audio and writes a one-line summary, so you can scan 20 responses in minutes. Built for customer discovery, sales outreach, user interviews, and stakeholder feedback.

Three steps. You write a question and create a Magic Link. You share the link. The recipient opens it in their browser, hits record, talks, and sends. On your side, every voice note arrives transcribed and summarized by AI. The whole loop, from sending the link to reading the answer, takes minutes.

No. The Magic Link opens in any browser on any device. There is no signup, no app to install, no login screen. The recipient taps record, talks for a minute, and hits send. The flow is built so the recipient never sees a friction point.

Yes. HeySpeak follows GDPR principles by default. We collect only the data we need, recipients are not tracked, no third-party cookies run on the recipient page, and every database table uses Supabase Row-Level Security. You can export or delete your data at any time. Full details on the data protection page.

Every layer sits inside the EU. Hosting runs on Vercel in Frankfurt. The database and authentication run on Supabase in Zurich. Audio files live on a private Cloudflare R2 bucket with EU jurisdiction. Voice transcription and AI summaries are handled by Mistral AI in France. No US-hosted subprocessor touches your recordings.

No. Voice transcripts and recordings are never used to train or improve AI models. We use Mistral AI, which contractually guarantees no training on API data. We do not sell data to third parties and we do not share recordings outside the subprocessor list.

Only you. Audio sits in a private bucket with no public URLs. To play a recording, your browser receives a 1-hour signed URL after Supabase Row-Level Security verifies you own the link the recording belongs to. The signed URL cannot be shared or cached beyond that window.

Yes. The free plan covers your first Magic Links and responses with no card needed, so you can run a real customer interview round before committing. Paid plans unlock higher limits and longer retention. See the pricing page for current numbers.

Use async voice when you want signal at scale. Twenty short voice notes in 48 hours teach you more than four 30-minute calls in two weeks. Use live calls for the two or three responses that surprise you. Async to filter, live to dig deeper.

Loom records you talking to your audience. HeySpeak records your audience talking to you, async and one-to-many. Typeform collects text without tone. Calendly books a synchronous call. HeySpeak gives you the depth of a conversation with the asynchronous reach of a form, plus an AI summary so you can scan all answers in minutes.

Warm contacts, yes. Most reply within a day, because asking for a minute of asynchronous voice is a much smaller ask than asking for a 30-minute slot in their calendar. Cold contacts respond less, around 20-30%. Send to 30 if you want 20 voice notes back.

Yes. Reply to the recipient with a second Magic Link and a follow-up question, or book a live call with the people whose first answer surprised you. The async-then-live sequence is usually faster than scheduling 20 live calls upfront.