Feedback for product teams

One tap tells you how it felt. Tunne finds out why.

Users tap a face at the moment it happens. An AI interviewer asks two or three honest follow-ups. Your team gets the cause, in the user's own words.

Try it: pick a face in the example and watch the follow-up.

How did exporting that report feel?

One tap. Then a few short questions.Powered by Tunne

Thanks. That's exactly what we needed.

The Northwind team reads every answer.

Moment: Report exported

Last 24 h · example

The problem

You have plenty of feedback. You don't know what it means.

  • Surveys arrive late.

    A monthly NPS mail asks about a moment the user forgot weeks ago. You get a summary, not the reaction.

  • Ratings have no why.

    A thumbs-down is a real signal you can't act on. Someone has to guess the cause.

  • Interviews don't scale.

    Research can talk to a handful of people a month. You ship every week.

  • Insight dies in a dashboard.

    Nobody opens the feedback tool on a Tuesday.

How it works

Here's what happens when someone uses your product.

You pick the moments to ask about, like a signup, a checkout or an export. Tunne does the rest.

  1. They finish an action

    You told Tunne this one matters.

  2. They tap a face

    One tap. Takes a second.

  3. An AI asks why

    Two or three short questions. It says it's an AI.

  4. You see the pattern

    Many answers become one clear problem, with real quotes.

  5. The right team gets it

    Sent to whoever owns that part of your product.

Emotion model

Furious and quietly disappointed are different problems.

Tunne places every reaction on two axes: how pleasant, and how charged. Each feeling opens the interview with a different question.

For your team

Evidence you can bring to a prioritisation meeting.

Every Moment in one place

Balance and charge, the themes behind them, and searchable verbatims — per Moment, not averaged into one score.

  • No NPS-style headline number
  • Delight counts as much as complaints
  • Quotes stay verbatim; summaries are labelled

Report exported · last 7 days · example

Delighted
38
Satisfied
61
Fine
24
Frustrated
17
Blocked
6

Balance +0.52 at mid charge — mostly working, one blocking theme.

Installs in a few lines

A script tag or the typed npm package, plus a mount point. Themed to look like your product, colours included. Asks each person at most once every 30 days, whichever surface asks.

<!-- Tunne capture widget -->
<script async src="https://cdn.tunne.ai/tunne.js"></script>
<div id="tunne-capture"></div>
<script>
  window.tunne("init", {
    publishableKey: "pk_live_…",
    mount: "#tunne-capture",
  });
</script>
// Or the typed package: npm install @tunne/widget
import { createTunneWidget } from "@tunne/widget";

const widget = createTunneWidget({
  publishableKey: "pk_live_…",
});
await widget.capture({ mount: "#tunne-capture" });

About

tunne, n. a feeling.
v. feel. Know.

In Finnish, tunne means an emotion. As a command it means feel — and also know. That's the product: capture the feeling, then come to know why.

  1. One tap is the whole commitment.Everything past the face is optional and has to earn the next second.
  2. The interviewer is honest.Disclosed as AI in the first turn. Short, non-leading, and stoppable.
  3. Say what the user said.Quotes beat paraphrase. Summaries are labelled as summaries.

Hear why, before everyone else.

Tunne opens to a small group first. Leave your email and we'll invite you.