What the person answering sees, and what you get back. Six questions takes them about four minutes; reviewing it takes you less than the call would have.
Plenty of people do their best thinking without an audience — on both sides of the link. If booking a call, making the small talk and being “on” for thirty minutes is the part of research you quietly dread, everything below happens over a link instead, at whatever hour suits the person answering.
Thirty seconds a question, counted down half a minute at a time as they go — not a hopeful “2 min” that turns into nine. A personalised link already knows who they are, so the contact form at the end disappears and the clock starts a minute lower. People finish what they can see the end of.
A text box is a filter. Most people write one flat sentence and move on. Hand the same person a microphone and they talk for forty seconds and give you the actual reason. No account, no app, no camera, nobody watching — they tap once and talk, alone in their kitchen if they like.
Answers arrive the way voice notes do: one thought, then the thing they remembered on the way to the kettle. They add a second take rather than starting over, so you keep the sentence that opens “oh, and honestly…” — the one a single-shot form throws away.
Every answer is written the moment it is given, never batched at submit. Someone who answers two questions on the train and never comes back is two answers and a timestamp, not an empty row. Partials sit next to finished responses, with the exact question where they stopped.
Waveforms are drawn when the answer is recorded, so you can see where someone got animated before you press play, and skip to it. Six people reviewed in the time one call would have taken — no diaries to align, no camera, no thirty minutes of being “on”.
Most people will open this on a phone, one-handed, somewhere with background noise. So the record button is thumb-sized and sits where your thumb already is, the answer saves the instant it is accepted, and the screen stays awake while the recording runs.
Most survey tools only show you what was submitted. If someone answers four of your six questions and closes the tab, that is four answers you never see.
Candid Reply writes every answer the moment it is given. Opening the link creates the response; accepting a recording saves it right then. The dashboard shows you who is at four of six, where they stopped, and what they said before they did.
Transcription starts on its own the moment a clip is uploaded. The audio stays the source of truth — a failed transcript never hides an answer, and the respondent never sees theirs, because reading it back turns a spoken answer into a written one.
Six answers from one person tells you about that person. Everyone’s answer to question three tells you about the question. A By question tab merges the transcripts into one block, labelled by who said what.
A search bar over every transcript in a survey. Both the question and the answer are saved with the number of responses it was based on, so the next person can read what was already asked.
CSV with links to the audio. Or an API key and an MCP connector, so Claude — or anything else that speaks MCP — can build surveys and read results without opening the app at all.
Unlimited surveys and questions on the free plan. No credit card.
Unlimited surveys and questions, free. No credit card.