In-app NPS survey questions

An in-app survey reaches people while they are using your product, so the answers come from active users with the experience fresh in mind. It also interrupts them, which is why the wording has to be shorter and the timing more careful than in an email.

Think of it as a small card in a corner: one line of question, a row of eleven numbers, and a text box that only appears after someone picks a score. Everything has to make sense at a glance and be easy to dismiss.

Questions to ask

How likely are you to recommend [Product] to a friend or colleague?
Slightly shorter than the classic wording, which helps on a small card. Do not shorten it further: "Would you recommend us?" invites a yes or no instead of a rating.
Sorry about that. What is getting in your way?
A short, human follow-up for low scores. In the product, "getting in your way" often points at a specific screen, which the respondent is probably looking at.
What one change would make [Product] better for you?
For the middle group. "One change" keeps the answer short enough for a small text box.
Thanks! What do you use [Product] for most?
For promoters. The answer tells you which job your best users hired the product for, which is valuable for positioning and onboarding.
Can we contact you about your answer?
A checkbox under the text box. Inside the product you already know who the person is, so only ask for permission, never for their email again.

When to show it

Show it after someone has used the product for a while and finished something, not the moment they log in and never in the middle of a task. Leave out brand-new users, people who answered recently and people who just dismissed it.

Respect a dismissal. If someone closes the card, do not show it again on the next page; wait until the next survey round. A survey that keeps coming back trains people to close it without reading.

Keep the card light

The survey should never slow the page down or cover the main navigation. Load it only when it is about to be shown, make it reachable by keyboard, and give every number a target large enough to tap on a phone.

How the answers are scored

Answers of 9 or 10 count as promoters and answers from 0 to 6 as detractors; the Net Promoter Score is the percentage of promoters minus the percentage of detractors.^{Source [1]} What counts as a good score depends on the industry and the survey method: see thecited NPS benchmarks.

Keep going

Answers to the follow-up question are often feature requests. On afeedback board they are counted with everyone else’s requests, and the people who asked hear when they ship.

Frequently asked questions

Is an in-app NPS survey better than an email survey?

It reaches active users at the right moment and usually gets more answers from them. Email reaches people who rarely log in, including those about to leave, so many teams use both.

Where in the app should the survey appear?

In a corner card or a slide-in panel that does not block the page, after the person finishes something. Avoid full-screen pop-ups and the first seconds of a session.

Should anonymous visitors see an in-app NPS survey?

Usually not. Ask signed-in users who have used the product enough to recommend it or not; a first-time visitor cannot answer the question honestly.

Sources

Bain & Company, “Measuring Your Net Promoter Score”: www.netpromotersystem.com/about/measuring-your-net-promoter-score/, retrieved 2026-10-25.

Net Promoter, NPS and Net Promoter Score are trademarks of Bain & Company, Inc., NICE Systems, Inc. and Fred Reichheld.

Ask these questions with UpVoiced

UpVoiced is adding NPS and CSAT microsurveys to its widget; they are coming soon and not available yet. Its feedback board, roadmap and changelog are in development, and you can try them on a fictional demo.

Check availability Try the fictional demo