Skip to content

Team reports

Team reports turn one survey into a report for every team: its scores beside the company and the team above it, its strengths and focus areas, and its comments when the team is big enough. Each team leader gets a private link to their team’s report. They don’t need an account. If you run the survey again, each report can also show how the team has changed since the earlier wave.

Team reports are available on the Growth plan and above. PDF downloads need the Team plan or above.

  1. Load your org structure: the teams, who leads them, and who belongs to each.
  2. Open the survey’s Settings → Data Retention → Team Reports and turn on Report by org unit. The survey takes its own copy of the org structure, and takes it again when it is first published.
  3. Send the survey. Each new response records its team:
    • By email: sent to the people in your org structure, each response records the invitee’s team.
    • By web link: add ?unit= and the team’s code to the link, for example ?unit=OPS-NORTH. A code that is not in the survey’s org structure records no team.

Once a survey has responses, Report by org unit can’t be turned off, because responses after that would have no team and every team’s results would be undercounted.

With Report by org unit on, a survey that collects responses anonymously still records no contact, IP address or device. It does record which team each response came from. That is the one thing it adds, and every view of it is held to the minimums below.

If you turn on Staff analysis access (below), the people who run the survey can also read responses from groups below those minimums. Each such view is recorded in the access log.

Team results follow two minimums, whether or not the survey is anonymous:

  • Scores need at least the survey’s Minimum group size of completed responses (5 unless you raise it).
  • Comments and individual responses need at least the Minimum group size for comments (10 unless you raise it; set it in Settings → Team Reports). Below it, a team’s charts and scores still show, but its text cards leave out recent answers, and the response list, text analysis, exports and AI features refuse a view of that team.

Both can be raised at any time, and neither can be lowered once the survey has responses.

Protection against working it out by subtraction. If one team under a department is below the minimum, the next-smallest team beside it is withheld too, so the department’s total minus the others never reveals the small team. If a withheld department has nobody directly in it, enough of its own teams are withheld as well that they can’t be added up to it: its smallest team, and then, by the same rule, the next-smallest beside it. Withheld teams still count in the department and the company totals.

No minimum is a guarantee. Two views that differ by one person, both above the minimum, can still isolate that person.

By default, nobody can see team results or comments below the minimums. If the people analysing the survey need response-level data for small groups, turn on Settings → Team Reports → Staff analysis access. Workspace owners, admins and editors can then read below the minimums, and every view they open below a minimum is recorded in the access log with who, when and what filter. Viewers, team leader links and the API stay bound by the minimums.

Dimensions are the scoring model of your survey, such as “Engagement” or “Wellbeing”. Each is a group of scale questions. Set them up on the survey’s Team reports page (open it from Settings → Team Reports → Open team reports):

  • Questions you can include: rating questions, NPS questions, slider statements, and the rows of a matrix question with single-choice answers (a Likert grid).
  • Scoring: Mean (the average of every answer to the dimension’s questions) or % favourable (the share of answers at or above a value you choose, such as 4 on a 1–5 scale).
  • Reversed: tick this for a negatively worded question, so a high answer counts as a low score. A matrix scale is scored by column position, first column = 1, so tick Reversed for a scale written best-first. The editor shows each scale’s first and last labels to help.

“N/A” answers are left out. A custom matrix column such as “Not sure” counts as a scale point, so leave such questions out of dimensions.

  • Each dimension and each question in a dimension: the team’s score, the company score (every response to the survey) and the score of the team above, with the difference from each.
  • Strengths and focus areas: the team’s highest and lowest questions, up to three each and never more than half the questions.
  • The questions where the team is furthest above and below the company.
  • Comments: answers to the survey’s open-text questions from the team and the teams under it, when the team meets the comment minimum.

A department’s report includes every team under it. A team that is withheld has no report. If the team above is withheld, the comparison with it is left out.

Preview any team’s report on the Team reports page before you send anything.

If you run the same survey again, for example once a year, each team’s report can show how the team has changed since the earlier survey.

On the survey’s Team reports page, under Earlier wave, choose the earlier survey in Compare with and select Save. The likeliest earlier waves, such as the survey this one was duplicated from, are listed first and marked “likely an earlier wave”. Only other surveys in the same workspace that report by org unit are offered. To choose one, you need to be able to edit this survey and see the team results of the earlier one.

Every team report, PDF and leader link then adds:

  • An Earlier wave and a Change column for each dimension and question.
  • Most improved and Most declined: up to three questions each, from the changes that are highlighted.

How the comparison works:

  • Questions are matched as in survey comparison: a question copied from the earlier survey matches its original even if it was reworded since, then questions with the same wording, and you can match the rest by hand with Match questions on the results dashboard.
  • The earlier survey is scored with this survey’s dimensions. Changing a dimension’s favourable value or its reversed questions doesn’t show up as a change.
  • Not compared: a question that was not asked on the same scale in the earlier survey shows “Not compared”, and so does every dimension that includes it.
  • Highlighting: a change is highlighted when it is too large for one person’s answers alone to explain: at least 5% of the scale, and more in a small team, for example 20% in a team of five. Smaller changes are shown, but not highlighted. The report says which threshold applies to the team.
  • Minimum group size: a team needs enough responses in the earlier survey too. If it had too few then, its report says so and shows no comparison.
  • Teams are matched by their team code in your org structure, so keep team codes the same between waves. A team with a code that wasn’t in the earlier survey has no comparison until you map it (see below). If most of the team codes in a roster file are new, the import warns you before anything is saved, so you can keep the old codes if the teams were only renumbered.
  • Response rate: the report gives the team’s response rate now and in the earlier survey, for example “Your team’s response rate: 48% now, 62% in Engagement 2025.” It’s shown even when the team had too few responses then for a comparison, because a response rate is a count, not a result. Each rate divides by the people the roster put in the team when the survey’s copy of your org structure was taken or last refreshed, so people who join or leave afterwards don’t change it. Surveys whose org structure was copied before 29 October 2026 have no recorded headcount, so they show no earlier rate.

Someone reading the report in SurveyRock sees the comparison only if they can see the team results of the earlier survey too. A team leader sees it on their link, because it is part of the report you send them.

When teams are split or merged between waves, the new teams have new codes and no earlier results. Under Earlier wave, Team mapping lists this survey’s teams that have no team with the same code in the earlier survey. For each one, choose the earlier teams it came from and select Save. Hold Ctrl (or Cmd) to choose more than one; choose none to clear a mapping.

  • Merged: a team that came from two or more earlier teams is compared with them taken together. Together they must reach the minimum group size, as a single team must. If any of them was withheld on its own (too few responses, or held back to protect a smaller team), the team shows no comparison, because the combined figures could reveal the small team’s.
  • Split: teams that came from the same earlier team are each compared with the whole of it, and the report says the team “was part of” it.
  • Renamed: a team mapped to one earlier team that no other team came from is compared with it, and the report says which team it was.

A mapping is kept for the earlier survey it was made for. If you choose a different earlier wave, map the teams again for that one. To map teams, you need the same access as to choose the earlier wave.

Duplicating a survey copies its dimensions. It does not copy its earlier-wave choice, so choose the earlier wave for each new survey.

On the Team reports page, Coverage lists each team: whether it gets a report, whether its comments are shown, and whether it has a leader email. It never shows how many people responded. For a team whose comments are shown, select Review comments and hide any comment you don’t want in team reports. A hidden comment is left out of that team’s report and every report above it. It stays in the survey’s results.

  1. On the Team reports page, under Leader links, choose how long the links work (90 days unless you change it, up to 365).
  2. Tick Email each leader their links if you want SurveyRock to send them.
  3. Select Make links. The page shows progress as it makes one private link per leader for each team that gets a report.

Each leader gets one email listing every team they lead. It comes from your workspace’s sender: your own mail server or verified sending domain if you have set one up. Unsubscribed and suppressed addresses are not sent to.

The email and the report are in the survey’s language: its headings, labels, numbers and dates, in any of the 21 languages SurveyRock’s survey messages come in, and right to left for a right-to-left survey. Questions and dimension names read as you wrote them.

A leader’s link opens their team’s report with no sign-in, on your custom domain if you have one. It shows your workspace’s logo and the survey’s own logo (Settings → Look & Feel), which you can use for your client’s logo. The report is checked against the minimums each time it is opened.

To stop every link in a batch, select Revoke all links next to it. Revoked links stop working at once.

On the Team plan and above:

  • A team’s report: Download PDF on the leader’s report, or in the preview on the Team reports page.
  • The whole survey: PDF on the Response Dashboard, or format: pdf in the API. It has a chart for every question, with the dashboard’s current filter, and no open-text answers. See PDF report.

Give your client’s HR contact a Viewer seat. A Viewer sees company-level results only: they can’t filter by team, and the Team reports page is not available to them. Owners, admins and editors see team results. On Enterprise, a custom role with Analyze set to Full does too.

Everyone who can see a survey’s results, Viewers included, has Response Rates by Team in the results sidebar. For each team it shows how many people the org structure puts in the team, how many have completed the survey, and the rate. It shows counts only, never answers, and is meant for chasing responses during fieldwork. “In the team” is the current org structure, not the list a particular email was sent to.

A team says “Withheld” but it has enough responses. It is being withheld to protect a smaller team beside it, or the team above it, from being worked out by subtraction.

A response has no team. It came from a web link without ?unit=, from someone who was not in the org structure, or before Report by org unit was turned on.

A new team is missing from the survey. The survey reports against its own copy of the org structure. Select Refresh from org structure in its settings.

A Viewer gets “Team-level results are not available to you”. That is the client HR restriction. Someone who should see team results needs an Editor seat or above.

Every team says there is no earlier result for it. Teams are matched by their team code, and the codes changed between the two waves. Map each team to the earlier teams it came from under Earlier wave → Team mapping (see Teams that were reorganised). Before the next wave, keep team codes the same if the teams themselves haven’t changed.