Back to blog

How to Create an NPS Survey: Question Design, Aggregation, and Connecting It to Improvement

与謝秀作

NPSアンケートの作り方|設問設計・集計方法・改善へのつなげ方

"I want to measure customer satisfaction with numbers, but I'm not sure what kind of survey to create." That's exactly when NPS (Net Promoter Score) comes in handy.

An NPS survey is appealing for how easily it measures customer loyalty from just one core question, but if you get the question design or aggregation method wrong, the hard-won data goes to waste. In this article, we explain how to create an NPS survey in three steps: question design, aggregation method, and connecting it to improvement.

What Is NPS? What an NPS Survey Tells You

NPS (Net Promoter Score) is a metric that quantifies how much customers want to recommend a product or service to others. It is widely adopted across many companies as an indicator of customer loyalty and intent to keep using a product or service.

Whereas a general customer satisfaction survey asks whether someone was satisfied, NPS is distinctive in asking whether they would recommend it to others. Because recommending is an action that signals support beyond mere satisfaction, it is said to correlate highly with future revenue and retention.

The main things an NPS survey lets you grasp are as follows.

  • The overall level of customer loyalty (the score)
  • The proportion of customers who will recommend you versus those at risk of churning
  • The reasons for recommendation or criticism (improvement hints gained from free-text responses)

Designing the NPS Survey Questions

An NPS survey is built around two questions: the core question and the reason question. Precisely because it is simple, the precision of how you ask determines the results.

The Core Question (Ask Recommendation Likelihood from 0 to 10)

At the heart of NPS is a question that asks recommendation likelihood on an 11-point scale. The standard way to ask it is as follows.

"How likely are you to recommend this product (service) to a friend or colleague? (0 = not at all likely to 10 = extremely likely)"

The key is to ask on the 11-point scale of 0 to 10. If you switch to a 5-point or 7-point scale, the basis for the aggregation described later breaks down, and you can no longer compare against other companies or past data.

The Reason Question (Free Text)

The score alone won't tell you why someone rated as they did. Place one free-text question asking for the reason right after the core question.

"Please tell us the reason for the score you gave."

This free-text response is the single most important source of information for improvement. Rather than chasing only the ups and downs of the score, pick up concrete issues from people's own words.

Three Things to Avoid in Question Design

  • Adding too many questions: It raises the response burden and invites mid-survey drop-off and lower answer quality. Keep it to a few questions: core, reason, and attributes.
  • Leading phrasing: Avoid expressions that prompt agreement, like "Were you satisfied?", and ask neutrally.
  • Changing the scale on your own: Anything other than 0 to 10 loses comparability. Stick to the standard.

How to Aggregate NPS (the Formula)

Once responses are in, classify respondents into three groups based on their recommendation score.

  • Promoters: 9-10 points. Loyal customers who actively recommend you.
  • Passives: 7-8 points. Satisfied, but without high enthusiasm.
  • Detractors: 0-6 points. Dissatisfied, and liable to churn or spread negative word of mouth.

NPS is calculated with the following formula.

NPS = percentage of promoters (%) - percentage of detractors (%)

For example, if out of 100 respondents there are 40 promoters, 35 passives, and 25 detractors, then 40% - 25% = 15 is your NPS. Note that passives are not included in the score calculation. The score ranges from -100 to +100.

How to Read the NPS Average and Benchmarks

Whether an NPS is good or bad varies greatly by industry, business custom, and country. Japan in particular tends to be reserved about recommending things to others, so scores are said to come out lower than in Western countries.

For that reason, getting elated or discouraged over the absolute value alone is a mistake. What matters is measuring continuously under the same conditions to track your own trend, and comparing it against benchmarks from peers in your industry.

Steps to Connect NPS to Improvement

NPS is not something you measure and forget; it gains value only when you run the improvement cycle. Combine the score with the free-text responses and proceed as follows.

  1. Analyze the voice of detractors: Read through the free-text responses from those who scored 0-6 and identify common threads in their dissatisfaction.
  2. Categorize issues and prioritize: Tackle frequently raised issues and those with the greatest revenue impact first.
  3. Execute improvements: Work with the relevant teams to decide whether to fix the product, support, or experience, and act.
  4. Also run measures to grow promoters: Reinforce the reasons for satisfaction and build mechanisms that encourage referrals and word of mouth.
  5. Measure again to verify the effect: Re-take the NPS regularly and confirm the effect of your measures and the trend.

Passives are especially easy to overlook. The 7-8 point passives have a high chance of turning into promoters with a small improvement in their experience, making them a segment with great upside.

Tips for Running an NPS Survey

  • Align the delivery timing: Deliver at a consistent timing that matches the experience you want to measure, such as right after purchase or after a support interaction.
  • Measure continuously: Making it a fixed-point observation rather than a one-off lets you see changes and the effects of your measures.
  • Don't leave responses unattended: A swift follow-up to low ratings can also be an opportunity to retain detractors.
  • Share the results with stakeholders: Share the score and customer voices with the front line to make improvement a company-wide effort.

Conclusion: Design Questions, Aggregation, and Improvement as One Continuous Flow

An NPS survey is fundamentally a very simple design: a core question asking recommendation likelihood from 0 to 10, and a free-text question asking the reason. For aggregation, subtract the percentage of detractors from the percentage of promoters to compute the score, and read averages and benchmarks through your own trend, taking the characteristics of your industry and country into account.

And the most important thing is to connect the score to action for improvement. By finding issues from the voice of detractors and running measures that push passives up to promoters, NPS becomes not just a number but a compass for nurturing your relationship with customers.

Back to blog