How to Run an NPS Survey in SurveyMonkey and Calculate Your Score

To run an NPS survey in SurveyMonkey, add the built-in Net Promoter Score question (or start from SurveyMonkey’s NPS template), send it to customers, and open Analyze Results: SurveyMonkey calculates the score automatically on a scale from −100 to 100. The score is the percentage of promoters (who answer 9 or 10) minus the percentage of detractors (0 to 6). What most guides leave out is how much the score can move by chance. With 200 responses, an NPS of 20 has a margin of error of about ±11 points, so a change from 20 to 25 next quarter may mean nothing. This guide covers the setup, how to read the results, and how to tell a real change from noise.

Disclosure: some links in this article may be affiliate links. If you sign up through them, scientyficworld.org may earn a commission at no extra cost to you. It does not change our recommendations.

What NPS Measures

Net Promoter Score asks one question: how likely are you to recommend us to a friend or colleague, on a scale of 0 to 10. According to the Net Promoter System site, respondents fall into three groups:

Group	Answer	Counted in the score?
Promoters	9 or 10	Yes, added
Passives	7 or 8	No, but they count in the total number of responses
Detractors	0 to 6	Yes, subtracted

NPS = % promoters − % detractors. If 45% of respondents are promoters and 25% are detractors, the NPS is 20. The score can range from −100 (everyone a detractor) to 100 (everyone a promoter).

How to Add an NPS Question in SurveyMonkey

These steps follow SurveyMonkey’s own help article:

1. Open your survey and go to Design Survey.
2. From the Build menu, choose the Net Promoter® Score question type.
3. Hover over the question and click Edit.
4. In the “this company” drop-down, choose an option or type your company, product, brand or service name.
5. Click Save.

You can change only the name of what is being recommended. The rest of the question wording and the 0 to 10 scale are fixed. That is deliberate: keeping the standard wording keeps your score comparable and makes the question eligible for SurveyMonkey’s benchmarks.

Or start from the NPS template

SurveyMonkey’s Net Promoter Score template opens with the NPS question, uses built-in skip logic to send each respondent to a follow-up open-ended question based on their score, and ends with demographic questions. The follow-up “why” is where most of the useful information is. Note that SurveyMonkey’s pricing page lists skip logic as a feature of paid plans, and its help centre confirms the free plan limits each survey to 10 questions, so check what your plan supports before relying on the template’s branching.

Reading Your Results

• The score: In Analyze Results, SurveyMonkey shows your NPS on a gauge chart from −100 to 100.
• The distribution: Switch to a bar or stacked bar chart to see detractors, passives and promoters, or the detailed distribution to see how many people chose each answer from 0 to 10.
• The data table: Below the chart, SurveyMonkey lists the exact count in each group. Use these counts, not just the percentage, when you report.
• Benchmarks: If you use the standard (“certified”) NPS question, SurveyMonkey benchmarks may be available, letting you compare your score with other organisations. The help article does not say which plans include benchmarks.

How Many Responses You Need: The Margin of Error

An NPS is an estimate from a sample, and small samples move a lot by chance. You can calculate a 95% margin of error from the share of promoters and detractors and the number of responses. Using the same mix (45% promoters, 30% passives, 25% detractors):

Responses	NPS	Margin of error (95%)	Plausible range
50	18	±22.6 points	about −5 to 41
200	20	±11.3 points	about 9 to 31
1,000	20	±5.0 points	about 15 to 25

With 50 responses, the true score could be anywhere from slightly negative to above 40. Before you celebrate or worry about a change between two surveys, check that the difference is bigger than the margin of error. As a rough guide, comparing two surveys combines the uncertainty of both, so with this mix of answers, reliably detecting a five-point change takes about 2,000 responses per survey, not dozens.

This short Python function calculates both the score and its margin of error from a list of raw 0 to 10 answers, such as a CSV export from SurveyMonkey. It reproduces the numbers in the table above:

from math import sqrt

def nps(scores):
    """scores: list of 0-10 answers. Returns NPS and its 95% margin of error, in points."""
    n = len(scores)
    promoters = sum(1 for s in scores if s >= 9) / n
    detractors = sum(1 for s in scores if s <= 6) / n
    score = (promoters - detractors) * 100
    variance = promoters + detractors - (promoters - detractors) ** 2
    margin = 1.96 * sqrt(variance / n) * 100
    return round(score, 1), round(margin, 1)

# Example: 200 responses -> 90 promoters, 60 passives, 50 detractors
example = [10] * 90 + [8] * 60 + [3] * 50
print(nps(example))
# Same proportions with 50 responses
print(nps([10] * 22 + [8] * 15 + [3] * 13))
# 1,000 responses, same proportions
print(nps([10] * 450 + [8] * 300 + [3] * 250))

Making NPS Useful for Product Teams

• Always ask why. The score tells you how many people would recommend you; the follow-up comments tell you what to change. Read every detractor comment.
• Segment before you average. Split the score by plan, customer age, region or feature use. A flat overall score can hide one segment improving and another collapsing.
• Track the trend, not the number. Survey the same kind of customers at the same point in their journey each time, and compare like with like.
• Close the loop. Reply to detractors where you can, and tell customers what changed because of their feedback.
• Don’t compare across contexts. A score from a post-support survey and one from an annual relationship survey are measuring different moments.

If you measure documentation or developer experience as well, our guide to measuring documentation success covers metrics that complement NPS.

Conclusion

SurveyMonkey makes the mechanics of NPS easy: a fixed, standard question, a template with follow-ups, and an automatic score in Analyze Results. The part that needs judgement is interpretation. Collect enough responses for the score to mean something, report the margin of error alongside it, segment the results and act on the comments. A score that moves within its margin of error is not news.