“What you see is all there is.” Daniel Kahneman · Thinking, Fast and Slow · 2011
Survey answers grouped into insights you can check.
Written answers, unsorted
3–5 days at checkout... it's day 9 and still nothing
Arrived a full week late, no update in the meantime
Tracking said Tuesday, showed up the following Monday
Wasn't sure what I'd get before creating an account
Signed up but couldn't tell what happens next
Asked for company details before showing a single screen
Wanted to try it, ended up filling out a form instead
Ten fields in before I saw any actual feature
Named insights, verified
- Shipping arrived later than promised34
"Said 3–5 days at checkout, took nine."
- Not clear what happens after sign-up21
"Wasn't sure what I'd get before creating an account."
- Setup asks too much before showing anything12
"Ten fields in before I saw any actual feature."
Hundreds of answers, and nobody reads them
How it works
One tap, wherever you put it
Embed the survey with a single script tag — your thank-you page, an onboarding step, a cancellation flow, any page you choose. The visitor answers one closed question with a single tap, saved the moment it happens, even if they close the tab straight after.
One short follow-up
Then one short written question, chosen by what they tapped — so you ask about the reason for that answer, not "what do you think". No email field, no discount — that is what the default template asks.
Counted honestly, not guessed
Once at least ten people have written something, matching answers are grouped and named with a sentence describing the problem. Below ten, you get a count instead of a guess — grouping a handful of answers produces something that looks convincing and means nothing.
See the answers behind every insight, not just the count.
Every insight opens into the answers it came from
An insight is a sentence describing a problem, with a count next to it. Open it and you get the replies it was counted from, word for word, no shortening and no paraphrase — and the list is exactly as long as the number says. The only text there that no respondent wrote is the insight's name, and it is labelled as generated.
Under ten answers, you get a count instead of an insight
Grouping a handful of answers produces something that looks convincing and means nothing. So below ten written answers this app shows no insights at all — it shows how many you have, how many it needs, how many are missing, and a link to the raw replies. Narrow a filter until you drop below ten and the insights disappear again, for the same reason.
When a recalculation fails, it says so
Insights are recalculated on a schedule. If a run fails, the panel says it failed and shows the last successful result with the date it was calculated — never a stale number pretending to be fresh, never an insight without a date. Your answers and filters keep working through it.
Two of those three are refusals, and that is the point. A tool that never says "I don't know" gives you no way to tell a real pattern from a filled screen. It also means there will be weeks when this app shows you nothing — we would rather do that than invent an insight.
Before anything reaches the AI
Sensitive details get masked before an answer reaches our AI provider
An e-mail, an IBAN, a phone number — respondents sometimes type them without thinking. On our own server, before an answer reaches our AI provider, we replace that kind of detail with a placeholder. The original stays in your account, and every quote you read in your panel comes from it — masking is best-effort and doesn't catch everything, but it helps you meet GDPR's data-minimisation principle for what leaves your account.
What leaves your account, and what the AI model sees, is in the privacy policy
Survey answer
Shipping was fast, but nobody replied when I emailed kontakt.klient@example.com about a discount code for order order-00048.
Pricing
Every plan runs on the same insights engine: quotes under every theme, and a refusal to count below ten answers.
- 100 responses per month
- 1 active survey
- 200 analyzed answers per month
- 1 user
- 5 recalculated insights per survey
- 90-day data retention
- Everything in Free, plus:
- 5,000 responses per month
- 5 active surveys
- 500 analyzed answers per month, up to 5,000
- 3 users
- 10 recalculated insights per survey
- 5 custom tracked insights (coming soon)
- Manual insight recompute, each run using analyzed answers from your monthly pool
- 360-day data retention
- Custom metadata on answers (coming soon)
- API access (coming soon)
- 14-day free trial
- Everything in Basic, plus:
- 100,000 responses per month
- 20 active surveys
- 1,000 analyzed answers per month, up to 50,000
- 5 users
- 10 recalculated insights per survey
- 20 custom tracked insights (coming soon)
- Manual insight recompute, each run using analyzed answers from your monthly pool
- 3-year data retention
- Custom metadata on answers (coming soon)
- API access (coming soon)
- MCP access (coming soon)
Common questions
How is this different from other survey apps that group answers with AI?
Grouping answers into insights is table stakes now — several apps do it, some at no cost. The difference is what happens around the insight: you can open the exact answers behind it, the app refuses to name any insight under ten written answers, and it tells you when a recalculation failed instead of showing you a stale result.
Why am I not seeing any insights?
Because there are fewer than ten written answers in the range you are looking at. You get a count instead — how many you have, how many are needed, how many are missing — never a guess.
Which sites does it work on?
Any page where you can add one script tag: your own site, a landing page, a web app, a documentation site. Some hosted checkouts do not allow scripts on their order confirmation page; there, put the survey on your other pages or send its link in the order e-mail.
Do you train AI on the answers you collect?
No. Answers are sent to a model provider for grouping only, and are never used to train or improve any model.
What happens if a recalculation fails?
The panel says the run failed and shows the last successful result together with the date it was calculated — never a stale number pretending to be fresh.
How much does it cost?
Free, Basic and Pro, from $0. Every plan is a superset of the one below it — see the full pricing page.