Product Signals

Product Signals is the feedback analytics dashboard in Trailguide Pro. Connect any app to send structured feedback responses — from in-trail feedback steps, support forms, or any other source — and get automatic sentiment analysis, categorical breakdowns, and word-frequency themes, all in one place.

How it works

1. Generate an API key in the Pro dashboard
2. POST feedback submissions from your app to the Trailguide API
3. View responses in the Product Signals dashboard at /dashboard/feedback

Each submission is a JSON object with a payload field that can contain any fields you want to track. Trailguide automatically detects sentiment scores, categorical breakdowns, and text themes from whatever you send.

Setup

await fetch('https://app.gettrailguide.com/api/feedback', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    'Authorization': `Bearer ${process.env.TRAILGUIDE_FEEDBACK_KEY}`,
  },
  body: JSON.stringify({
    formId: 'contact-import-feedback',     // identifies this form in the dashboard
    formLabel: 'Contact Import Feedback',  // human-readable label (optional)
    appId: 'my-app',                       // your app identifier
    instance: 'acme-corp',                 // tenant / subdomain / customer ID
    respondentName: 'Jane Smith',
    respondentEmail: 'jane@acme.com',
    payload: {
      // any fields you want to track
      source: 'Google Contacts',
      rating: 4,
      comment: 'Import was smooth but the field mapping could be clearer.',
    },
  }),
})

Request body fields

What the dashboard shows

The Product Signals dashboard has three levels of detail:

L1 — Snapshot bar Three at-a-glance metrics with trend arrows comparing the current period to the previous one:

• Total submissions and change vs. prior period
• Average sentiment — automatically detected from any numeric rating field in your payload (1–5 or 1–10 scale, or fields named rating, score, nps, stars)
• Top word — the most frequent meaningful word across all text responses

L2 — Breakdown charts

• Sentiment distribution — histogram of rating values if a sentiment field is detected
• Categorical breakdowns — for any payload field with 2–15 distinct string values (like source, plan, role), a bar chart shows submission counts per category
• Word themes — top keywords extracted from free-text fields (fields where the average response length is over 30 characters), with stopwords filtered out

L3 — Raw submissions table (collapsed by default) Every individual submission with all payload fields, respondent, instance, and timestamp.

Time period selector: 7d / 30d / 90d / 6mo / 1yr / All — switching is instant.

Backfilling from Jira

If your team previously collected feedback as Jira comments (in the format Instance: X\nUser: Y (email)\nSource: Z\n...), you can import them in one click. In the left panel next to any form, click Sync from Jira. Trailguide fetches the Jira issue’s comment history, parses each comment into structured fields, and imports them as submissions with sync_source: jira_import.

Using with in-trail feedback steps

If you use the feedback step type in a trail, you can forward those responses to Product Signals by configuring the step’s webhook field to point at your own server, then forwarding the payload to the Trailguide API:

// Your server — receives the webhook from the feedback step
app.post('/feedback-webhook', async (req, res) => {
  const { rating, comment, trailId, stepId } = req.body
 
  await fetch('https://app.gettrailguide.com/api/feedback', {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
      'Authorization': `Bearer ${process.env.TRAILGUIDE_FEEDBACK_KEY}`,
    },
    body: JSON.stringify({
      formId: trailId,
      appId: 'my-app',
      instance: req.user.tenantId,
      respondentName: req.user.name,
      respondentEmail: req.user.email,
      payload: { rating, comment, stepId },
    }),
  })
 
  res.sendStatus(200)
})

This connects the user experience of the tour with the analytics of their response.

CORS

The POST /api/feedback endpoint accepts requests from any origin (Access-Control-Allow-Origin: *). You can call it directly from the browser if your API key is acceptable to expose in client code, though posting from your server is recommended.