Continuity
Keeping your engagement scores comparable across a platform change
Why a score moves when the tool changes, what still compares, and what to say before the first result lands.
4.7/5 on G2
SOC 2 Type II ISO 27001 GDPR
Engagement programs that act, not just measure
Show your working
The Help Center’s own example: three drivers, 500 rating answers.
Recognition3.20
Growth & Development3.00
Wellness4.40
Overall3.88
Open text · separate reporteNPS · separate report
Stored 1 to 5Displayed7.8 / 10
Mechanism
Why scores move when the platform moves
- Culture Amp reports a percentage. CultureMonkey reports an average. Culture Amp scores engagement as the average percent favorable: the share of people who picked Agree or Strongly agree, the 4s and 5s on its five-point scale. CultureMonkey averages every answer instead, holding ratings on 1 to 5 and showing them on 0 to 10 by default. One counts how many people cleared a bar. The other averages where everyone stood.
- Which is why no conversion factor exists. Take two teams that both average 3.0 out of 5. The first splits evenly between 5s and 1s; the second answered 3 straight down. Both read 6.0 on CultureMonkey. On Culture Amp the first is 50 percent favorable and the second is 0. Identical averages, fifty points apart, and neither platform is wrong.
- Changing the display is the easy half. A CultureMonkey rating simply doubles for the 0 to 10 view, so 3.88 out of 5 shows as 7.8, and the same figure can be shown out of 5 or as a percentage instead. Nothing performs the equivalent trick on a percent favorable, because the people who answered 3 were dropped from it before the number was formed.
- Wording and question count move the score on top of all that. Asking twenty questions rather than forty changes which opinions are averaged. Rewording one changes what people answered.
Worth knowingThis happens with any platform change, in any direction. It is not a CultureMonkey quirk or a Culture Amp failing. Knowing it early lets you explain the number, not defend it.
CultureMonkey benchmark · 2025 to 2026How much engagement really moves in a year
+0.06All CultureMonkey customers, 0 to 10 scale
100 to 500 employees7.93 → 8.10+0.17
500 to 2,000 employees8.22 → 8.24+0.02
Questions
Bringing your questionnaire across without losing comparability
- Keep your existing wording; it is usually worth it. Identical words remove one of the three causes above, so any movement is easier to explain to a board.
- Each rating question is re-created and mapped to a driver. That mapping is required, not optional, so allow setup time for a long survey.
- Standard drivers can have a benchmark; custom drivers cannot. Library questions on standard drivers appear in benchmark comparisons where your industry has data.
The quiet trapBenchmarks are tied to drivers, not your exact wording. Heavily reword a library question and the comparison is no longer fair, though it still looks valid on screen.
Before you rebuild anything, check what you need to export first.
Driver set · survey settings
2 modesA survey uses one set or the other, never a mix.
SummaryReportQuestionsBenchmark
Standard drivers compare against your industry and employee-size band on the Benchmark tab.
Q2 board recap
Engagement
Up 6 points this quarter
42eNPS
88%Participation
17/20Actions
Export deckPDFPPT🔗 Link
Board
What to tell your board before the first result
- Say the number will move before you know which way. Predicted, it reads as a measurement change. Announced afterward, it reads as your result.
- Name the three reasons: scale, wording, count. One sentence each. Executives accept a measurement change when the cause is specific, not when it sounds like reassurance.
- Pick the cycle you start counting from, and say so. Report movement from that point. Keep earlier scores as history, not quietly blended with new ones.
Copy into an email
A version you can paste into an email: we are moving engagement surveys to a new platform this quarter. The headline score will change because the rating scale, wording and number of questions all differ. That reflects how we measure, not how people feel. We will report movement from this cycle on, keep previous scores on record, and track the drivers, where our actions come from.The timing of that first result depends on when in your survey cycle to switch.
FAQ
Engagement scores across a platform change: common questions
Will our engagement score drop when we change survey platforms?
It will almost certainly move, up or down, usually for reasons unrelated to your organization. Scale, wording and question count differ between platforms, and each one shifts the overall score.
Can you compare engagement scores across two different survey tools?
Not as raw numbers, because platforms rarely share a scale, wording or weighting. What does compare is direction and relative standing: which drivers lead, which lag, and which way each moves.
How does a Culture Amp favorability score convert to a CultureMonkey score?
It does not convert, because the two are different measurements rather than two scales of one measurement. Culture Amp reports the average percent favorable, the share of people choosing Agree or Strongly agree on its five-point scale. CultureMonkey averages every answer, holding ratings on 1 to 5 and showing them on 0 to 10. Two teams that both average 3.0 out of 5 read 6.0 in CultureMonkey, while one can be 50 percent favorable and the other 0 percent in Culture Amp. Report movement from your first CultureMonkey cycle and keep the Culture Amp figures as history.
Why are CultureMonkey ratings stored on 1 to 5 but shown on 0 to 10?
Ratings are stored 1 to 5 and shown 0 to 10 by default, or out of 5 or as a percentage. That display choice alone changes the number people see, so an old score is not a fair comparison.
Will our benchmark comparison carry across from Culture Amp?
No, and it is not meant to. Benchmarks are tied to drivers, not to a vendor. Library questions on standard drivers appear in benchmark comparisons. Custom drivers do not, since only you measure them.
Keep your engagement scores comparable when you switch
We will map your questions to drivers live and show which can have a benchmark, before you commit to anything.
4.7 / 5 on G2
This page is part of the full Culture Amp migration guide. Still deciding? Our roundup of Culture Amp alternatives compares the field, and this page applies whichever you pick.