What you do in the four weeks after an engagement survey determines whether anyone answers the next one honestly. The most common failure in employee listening is not survey design or response rates — it is the silence that follows.
This is a dated thirty-day plan. It is deliberately specific about timing, because the credibility of the whole exercise is decided by how quickly people hear something back.
Why the first 30 days matter more than the survey itself
Employees form a judgement about whether feedback is worth giving based on what happened last time. If a previous survey produced no visible change, response rates fall and the answers that do come back get safer.
The reverse is also true and considerably underused: a single visible change, clearly attributed to survey feedback, does more for the next cycle’s response rate than any amount of reminder emails.
The most important sentence in your post-survey communication is not what you are changing. It is “you told us this, so we are doing that.” The attribution is what teaches people that answering works.
The 30-day plan
| When | What happens | Who |
|---|---|---|
| Days 1–3 | Thank everyone. Share the response rate. | HR |
| Days 4–7 | Analyse. Segment by site, team, tenure. Theme the comments. | HR |
| Days 8–10 | Managers receive their own team results first. | HR → managers |
| Days 11–14 | Share organisation-wide results openly, including the bad parts. | Leadership |
| Days 15–21 | Managers hold team conversations about their own results. | Managers |
| Days 22–28 | Agree actions — two or three, with owners and dates. | Managers + HR |
| Day 30 | Communicate what is changing and why. | Leadership |
| Week 6–8 | Pulse the specific issue to see whether it worked. | HR |
Days 1–3: Say thank you, and say the number
Within 72 hours of close, tell everyone the survey has closed and what the response rate was. Nothing else is needed yet.
This is a small thing that signals something is happening. The alternative — complete silence for three weeks while analysis takes place — is where the assumption that nothing will change begins.
Days 4–7: Analyse properly
Three things to do, in order.
Segment before you read the headline. An organisation-wide score of 72% is nearly useless on its own. A range of 61% to 88% across twelve sites is a set of specific conversations. Segment by site, team, role, tenure and employment type.
Read the comments, or theme them. Scores tell you what is happening; comments tell you why. At scale, manual reading is impractical — 500 employees writing a paragraph each takes weeks. AI theming groups them by topic and sentiment in hours.
Identify the gap, not just the low score. Look for the driver with the biggest gap between how much it influences engagement in your data and how it currently scores. The lowest-scoring driver is not always the one worth acting on.
Days 8–10: Managers see their results first
This step is skipped more often than any other and it matters disproportionately.
A manager who first hears about their team’s recognition score in a leadership meeting will be defensive. A manager who saw it a week earlier has had time to think about what to do. Give them their own team’s results, their strongest and weakest drivers, and the comments relevant to them — subject to minimum reporting thresholds so small teams stay confidential.
Prosper Employee Engagement Give every manager their own team results Prosper reports engagement by manager, team and site, with confidentiality thresholds built in. Managers see their weakest driver and the comments that explain it — which is what turns engagement from an HR metric into something they can act on. See Prosper Employee EngagementDays 11–14: Share results openly, including the bad parts
Share the whole picture, not a curated version. Employees generally know what the difficult findings are — they wrote them. A leadership summary that omits them signals that the difficult things will not be addressed.
Include the response rate, the highest and lowest scoring drivers, the movement since last cycle, and the main comment themes. Name the things that are not good.
Days 15–21: Managers hold team conversations
Not a presentation. A conversation, in each team, about that team’s own results.
A structure that works: here is what our team said, here is what I think it means, here is what I want to ask you about, here is what I can change and what I cannot. The last part matters — being honest about constraints is more credible than promising everything.
Days 22–28: Agree two or three actions
Not twelve. Two or three, each with a named owner and a date.
| Weak action | Strong action |
|---|---|
| “Improve communication” | “Roster published 3 weeks ahead from 1 September — owner: Ops” |
| “Focus on recognition” | “Every manager names one person in Monday brief — from next week” |
| “Better development opportunities” | “Two supervisors from each venue into the leadership programme by Q4” |
| “Address workload concerns” | “Review Friday rosters at the three sites that raised it — by 30 Sept” |
Most action should be owned at site level rather than centrally. The problems are usually local, and a central action plan produces central activity rather than local change.
Day 30: Communicate what is changing
Short, specific, and attributed. “You told us rosters were being published too late. From September they will be out three weeks ahead.”
Include what you are not doing and why. Employees accept a reasoned no far better than silence — and silence gets interpreted as a no anyway, just with the reasoning invented.
Weeks 6–8: Measure whether it worked
The step almost nobody takes. Run a short pulse survey on the specific issue you acted on. Did the score move at the sites where you made the change?
This does two things: it tells you whether the action worked, and it demonstrates that feedback is genuinely part of a loop rather than an annual event. Our guide to building a listening strategy covers how this becomes a permanent cadence.
What to do when the results are bad
Three rules for a poor result.
Do not delay the communication. The instinct is to hold results until there is a plan. Employees interpret the delay as concealment, and the plan arrives too late to be credible.
Do not explain it away. “Scores were affected by the restructure” may be true and it still reads as an excuse. Acknowledge first, contextualise second, if at all.
Pick one thing and fix it visibly. A single completed change beats a comprehensive plan. Credibility is rebuilt through evidence, not intention.
Building a twelve-point action plan after a poor result. It signals seriousness to the executive team and produces nothing at site level, because twelve priorities is no priorities. Two actions, completed and communicated, will do more for the next cycle’s scores than a comprehensive plan that stalls in month three.
Common mistakes
Silence. The single biggest one. Weeks of analysis with no communication teaches people that nothing is happening.
Sharing only the good findings. Employees wrote the bad ones. Omitting them destroys trust in the whole exercise.
Central ownership of local problems. HR cannot fix a rostering issue at one venue. The venue manager can.
No attribution. Changing something without saying it came from the survey wastes most of the benefit.
Never re-measuring. Without a follow-up pulse, you never learn whether the action worked, and neither does anyone else.
See it in practice Close the loop, then prove it closed Run a targeted pulse six weeks after acting and see whether the score moved at the sites where you made the change — segmented by team, role and tenure. Explore Prosper Employee EngagementFrequently asked questions
Thank people and share the response rate within 72 hours, analyse and segment within a week, give managers their own team results before the organisation-wide presentation, share results openly including the difficult findings, agree two or three actions with owners and dates, and communicate what is changing by day 30.
Acknowledge the survey has closed within 72 hours, and share full results within two weeks. Longer delays are interpreted as concealment, and the eventual communication arrives too late to be credible.
Yes. Employees wrote the difficult findings and generally know what they are. A curated summary that omits them signals the difficult things will not be addressed.
Two or three, each with a named owner and a date. Twelve priorities is no priorities, and comprehensive action plans typically stall by month three.
Mostly site or team managers rather than HR centrally. The problems are usually local, and central ownership produces central activity rather than local change.
Specificity. "Improve communication" is not an action. "Rosters published three weeks ahead from 1 September, owned by Ops" is.
Run a short pulse survey on the specific issue six to eight weeks later and check whether the score moved at the sites where you made the change. Most organisations skip this step and never learn.
Say so, and say why. Employees accept a reasoned no far better than silence — and silence gets interpreted as a no anyway, with the reasoning invented for you.
Close the loop — then prove it closed
Run a targeted pulse six weeks after acting and see whether the score moved at the sites where you made the change.