How to run an NPS survey

Decide what the NPS survey is for

Before you send an NPS survey, decide which decision it supports. A relationship NPS survey goes to your customer base at a steady interval and tracks loyalty to the brand. A transactional NPS survey goes after a specific interaction and tells you how that touchpoint affects loyalty. Both use the same question; they answer different questions for the business.

Use the standard question

The recommend question comes from Frederick Reichheld's 2003 Harvard Business Review article. The usual wording is "How likely are you to recommend [brand] to a friend or colleague?" on a 0 to 10 scale, with 0 labelled "not at all likely" and 10 "extremely likely". Keep the wording and the scale exactly the same every time; small changes break comparison.

Choose who to ask

• Relationship survey: active customers who have bought or visited in a recent window, sampled fairly across regions and segments.
• Transactional survey: customers who just completed the interaction you are measuring.
• Exclude staff, test accounts and customers surveyed very recently.
• Set a contact rule, for example no more than one NPS invitation per customer per quarter.

Pick the timing

For transactional NPS, send within a day of the interaction while it is fresh. For relationship NPS, pick a cadence you can sustain, often quarterly or twice a year, and send at a similar point each time so seasonal effects do not masquerade as change.

Choose the channel

Use the channel your customers already use with you: email for online customers, SMS or WhatsApp where you have consent and that is how you communicate, a QR code on a receipt for walk-in customers. Keep the survey short enough to finish on a phone.

Add the follow-up question

The score alone does not tell you what to do. Add one open question straight after it: "What is the main reason for your score?". Optionally, ask a different second question by group: "What would make you more likely to recommend us?" for passives and detractors, "What do you value most?" for promoters.

Calculate the score

Group answers into promoters (9 or 10), passives (7 or 8) and detractors (0 to 6). NPS is the percentage of promoters minus the percentage of detractors, so it runs from -100 to +100. For example, 120 promoters, 50 passives and 30 detractors out of 200 gives 60% minus 15%, an NPS of 45.

Watch the sample

Report the number of responses next to every NPS. A store with 25 responses can swing many points because of a handful of customers. Use longer windows for small locations, and do not rank stores on tiny samples.

Analyse the comments

Code the open answers into themes (staff, speed, price, product quality, cleanliness, app) and compare theme frequency between promoters and detractors. The themes that appear mostly in detractor comments are your priorities; the ones that dominate promoter comments are what to protect.

Close the loop

• Contact detractors who left a way to reach them, within an agreed time.
• Route specific complaints to the store or team that owns them.
• Share recurring themes with leadership with quotes, not just the number.
• Tell customers and staff what changed as a result.

Keep it consistent

Write down the question, scale, audience, timing, channel and calculation in one short method note. Changes to any of these can move the score without any real change in loyalty. If you must change something, run old and new in parallel for a period and note the break in your charts.

Common mistakes

• Changing the question wording or scale.
• Surveying only happy customers, such as loyalty members.
• Reporting NPS without response counts.
• Skipping the open follow-up.
• Chasing the score with staff incentives, which invites score begging.

Common questions

Checklist

• Survey type chosen.
• Standard question and 0 to 10 scale.
• Audience and contact rule set.
• Timing and channel fixed.
• Open follow-up included.
• Score calculated and shown with counts.
• Comments themed and acted on.

In short

To run an NPS survey, choose relationship or transactional, ask the standard 0 to 10 recommend question with one open follow-up, send it at a consistent time and channel, calculate promoters minus detractors, and act on the themes in the comments.

Doing this with Falcon

Falcon does not run surveys or feedback forms. It analyses the unprompted feedback customers already leave in Google Maps reviews of your locations and Google Play reviews of your app: each review is classified into themes, sub-themes and specific issues from your own taxonomy, with the sentiment of each mention and the exact quote behind it. Detractor comments and negative reviews often name the same problems. Falcon shows negative share by theme, store and month from reviews, which you can line up against the themes in your NPS comments to check whether both point the same way.