After publishing an NPS survey, track responses, monitor your score over time, and export data for reporting. NPS 2.0 analytics are split across two tabs — Overview for engagement metrics and NPS for score data and individual responses.

Open NPS analytics

1. Go to Experiences > NPS and click on your survey. The analytics page opens.
2. Use the date range picker to set the reporting period. Use the segment filter to view data for a specific audience.

Overview tab

The Overview tab shows engagement metrics — how many users saw the survey and how many completed it. This uses the same analytics format as Flows.

Flow activity chart

The chart plots two lines over your selected date range: Flows shown (how many times the survey was displayed) and Flow completion (how many users completed the full survey). Use this to spot trends — a spike in "shown" with flat "completion" may indicate users are dismissing or snoozing the survey.

Summary cards

• Unique Users — the number of distinct users who were shown the survey, plus the total number of times it was shown (one user may see it multiple times across cycles). For example, "2,784" unique users with "13,813 total times shown."
• Unique Users Completed — the percentage of unique users who completed the survey (submitted at least a score), plus the overall completion rate across all impressions.

Step breakdown

The step breakdown table shows per-step metrics for each step in the survey:

Step 1 is the score question. Step 2 is the feedback question. A high completion rate on Step 1 but a drop on Step 2 means users are submitting scores but skipping the feedback question — this is normal, but if the drop is steep, consider shortening or simplifying your feedback prompt.

Recent Users

A table of the most recent users who interacted with the survey, showing User (clickable — opens the user's profile) and Event Time.

NPS tab

The NPS tab shows your actual NPS score, the promoter/passive/detractor breakdown, score trends, and individual responses.

Score tile

The score tile shows your current NPS score and the response breakdown:

• Score — your NPS, calculated as % promoters minus % detractors. Ranges from −100 to +100.
• Stacked bar — a horizontal bar showing the percentage split: promoters (green), passives (grey), and detractors (red).
• Counts — the number of respondents in each category (e.g., 119 Promoters, 41 Passives, 123 Detractors).

For example, if you have 1,200 people respond to your NPS survey and 400 are promoters and 200 are detractors, your NPS score would be 17. 

[(400/1,200) - (200/1,200)] x 100 = 17

The score is calculated using a trailing 30-day window from the end of your selected date range, so it stays current and smooths out day-to-day spikes.

Score history

A line chart showing your NPS score over time across the selected date range. Each data point represents the trailing 30-day score calculated on that date. Use this to track whether your score is trending up, down, or holding steady.

For example, a score calculated on March 5 uses data from February 4 to March 5. The next day, it shifts to February 5 to March 6.

Survey Responses

A table of individual responses with four columns:

If a user submits multiple scores within the same cycle (e.g., they updated their score), both responses appear in the response table but only the most recent score is used in the NPS calculation.

Export NPS data

Click the Export button in the top-right corner of the analytics page. Select a date range, choose an export type, and a CSV will be generated and sent to your email.

Export types

• Basic — Includes core activity events for the NPS experience (appcues:v2:experience_started, appcues:v2:step_seen, appcues:v2:experience_snoozed, appcues:v2:experience_completed) with reserved fields like timestamp, user ID, and event name. It does not include score or feedback. Use this for a lightweight view of who interacted with the survey and what happened.
• Full — Same as Basic but expanded to include all event attributes. Every user property and attribute tied to each event gets its own column. This can result in very wide CSVs. Use this when you need the most comprehensive raw event dump.
• Survey — One row per respondent. Combines the score and feedback responses for a single user into a single row, so you get their NPS score and text feedback together alongside user properties. Only includes users who actually responded — not users who were shown the survey but didn't submit. This is the most useful export for most teams doing NPS analysis.

The Survey export is usually what you want. Basic and Full are row-per-event formats designed for piping into analytics tools or data warehouses. Survey gives you a clean, one-row-per-person view that's ready for spreadsheet analysis.

NPS 2.0 event model

NPS v1 used dedicated event names like appcues:nps_score and appcues:nps_feedback. NPS 2.0 eliminates those in favor of a generic, extensible survey event model appcues:v2:step_interaction.

Event structure

All NPS 2.0 responses come through as appcues:v2:step_interaction events with interactionType set to "Survey submitted". The NPS data lives nested inside interactionData:

{
  "name": "appcues:v2:step_interaction",
  "experienceId": "...",
  "experienceName": "NPS for legacy customers",
  "experienceType": "nps",
  "stepId": "...",
  "interactionType": "Survey submitted",
  "interactionData": {
    "survey": {
      "id": "<survey-entity-uuid>",
      "type": "nps"
    },
    "submissionId": "<unique-submission-uuid>",
    "surveyResponses": [
      {
        "type": "nps_score",
        "label": "NPS Score",
        "value": "9",
        "fieldId": "<survey-field-uuid>",
        "order": 0
      },
      {
        "type": "nps_feedback",
        "label": "NPS Feedback",
        "value": "this is awesome",
        "fieldId": "<survey-field-uuid>",
        "order": 1
      }
    ]
  }
}

Key fields

• survey.id and survey.type tie the response back to a specific survey entity. This is what enables multiple NPS surveys — each has its own survey entity ID.
• surveyResponses is an array where each entry has a type (like nps_score or nps_feedback), a fieldId linking back to the survey entity's field definition, a label, and a value.
• submissionId uniquely identifies this particular submission.
• Score and feedback can arrive in separate submissions — a user might submit a score in one event and follow up with feedback in another.

Other NPS 2.0 events

Survey seen events (for calculating response rate) also use appcues:v2:step_interaction but with interactionType: "Survey seen" and only the survey object in interactionData — no surveyResponses.

Snooze events map to appcues:v2:experience_snoozed, replacing the old appcues:nps_ask_me_later_selected_at.

V1 → V2 event mapping

Integration support for NPS 2.0

Good to know

• Score calculation. Your NPS = % promoters − % detractors. Passives are excluded. Appcues uses a trailing 30-day window so the score stays current.
• Multiple responses. If a user submits multiple scores within a cycle, only the most recent score is used in the NPS calculation. Both responses appear in the Survey Responses table.
• Segment filtering. Use the segment dropdown to see how your NPS varies across different user groups — by plan, role, tenure, or any segment you've created.

If the data doesn't look right

• Score seems off? Remember the score uses a trailing 30-day window, not just the selected date range. A short date range may show fewer responses than expected.
• Responses missing? Check that the survey is published and that users are meeting the targeting and sampling criteria. See NPS targeting settings.
• Export is empty? Verify the date range includes dates when the survey was live and receiving responses.
• Integration not receiving NPS 2.0 events? V2 uses appcues:v2:step_interaction instead of appcues:nps_score. We're still rolling out support to NPS 2.0 across our integrations.