Metrics

Three endpoints return the numbers the dashboard shows. They need a credential with the analytics:read scope.

Endpoint	Returns
`GET /metrics/summary`	The response count, and for one survey its NPS, CSAT, CES and average score
`GET /metrics/timeseries`	One value per day, week or month
`GET /metrics/breakdown`	Responses per value of one dimension

All three take surveyId, from and to. A to given as a date alone, such as 2026-10-14, includes that whole UTC day. Without surveyId, the numbers cover the whole organization. Without from and to, they cover everything your plan keeps. A from older than your plan’s retention window starts at the beginning of the window instead.

Get the summary

curl "https://app.usertick.com/api/public/v1/metrics/summary?surveyId=7c9e6679-7425-40de-944b-e07fc1f90ae7&from=2026-09-01T00:00:00Z" \
  -H "Authorization: Bearer utk_live_your_key_here"

{
  "surveyId": "7c9e6679-7425-40de-944b-e07fc1f90ae7",
  "from": "2026-09-01T00:00:00.000Z",
  "to": null,
  "responses": 412,
  "nps": { "value": 31.443298969072163, "sampleSize": 388, "questionId": "4b1c2d3e-0000-4a5b-8c9d-0e1f2a3b4c5d" },
  "csat": null,
  "ces": null,
  "avgScore": null
}

Each score is read from the survey’s first question of its kind, the way a number card on a dashboard reads it. A score is null without surveyId, and when the survey has no question of that kind. Scores are unrounded. NPS runs from -100 to 100, CSAT is the percent of ratings at or above the satisfied rating (4 of 5, 8 of 10), and CES and the average score are the average answer. Round each value once where you show it: the dashboard shows the NPS above as +31 and an average to one decimal.

Get a time series

metric is responses (the default), nps or detractors. interval is day (the default), week or month, in UTC. The nps and detractors metrics need a surveyId.

curl "https://app.usertick.com/api/public/v1/metrics/timeseries?metric=nps&interval=week&surveyId=7c9e6679-7425-40de-944b-e07fc1f90ae7" \
  -H "Authorization: Bearer utk_live_your_key_here"

Each point is { date, value, count }, oldest first. For nps, value is the step’s NPS, unrounded, and 0 when count is 0; treat a step with no answers as having no value.

Get a breakdown

by sets the dimension:

`by`	Counts responses per
`type`	Feedback type: bug, idea, complaint, question, praise
`urgency`	Urgency: high, medium, low
`theme`	Theme in your analysis tree
`category`	Category in your analysis tree
`sentiment`	Sentiment: negative, neutral, positive
`country`	Country, with `unknown` and `other` buckets
`device`	Device type, with `unknown` and `other` buckets

For country and device, top sets how many values come back before the rest are counted under other. It takes 1 to 250 and defaults to 12. A value outside that range returns 400 with PUB-006-005. The analysis dimensions always return every value.

{
  "by": "type",
  "surveyId": null,
  "total": 1280,
  "buckets": [
    { "key": "bug", "label": "Bug", "count": 402 },
    { "key": "idea", "label": "Idea", "count": 311 }
  ]
}

Buckets are sorted largest first. The analysis dimensions include corrections people made. A response with several answers can count under more than one value, so the buckets can add up to more than total. A response with no value on the dimension, such as one not analyzed yet or one that matched no theme, has no bucket, so the buckets can also add up to less than total. To list the responses behind a bucket, call GET /responses with the same filter, such as type=bug.

JavaScript SDK

REST API

API reference

Webhooks

MCP

Get the summary

Get a time series

Get a breakdown

​Get the summary

​Get a time series

​Get a breakdown

Get the summary

Get a time series

Get a breakdown