You finally added a survey to your app. A week later you check the numbers: 1,000 people saw it, 60 answered. That’s 6%. Is that bad?

Probably, yes. But the number on its own tells you less than you think. Before you start rewriting questions, it helps to know what the benchmarks actually measure, what moves the number, and when a higher response rate doesn’t mean better data.

The quick answer

For surveys shown inside a mobile app, published benchmarks land roughly between 13% and 36%:

SourceWhat was measuredAverage response rate
Alchemer Mobile (formerly Apptentive)In-app surveys across mobile app categories13%
Refiner, 2025 report1,382 in-app surveys, 50 million views, web and mobile27.5% overall
Refiner, 2025 reportMobile app surveys only36.1%
Refiner, 2025 reportNPS surveys / CSAT surveys21.7% / 26.3%

Refiner also publishes a simple scale for mobile apps: above 19% is good, above 35% is great, and above 62% is excellent.

So if you’re below 10% in a mobile app, something structural is probably wrong. If you’re between 15% and 35%, you’re in normal territory and the next gains come from smaller changes.

One honest caveat: all of these benchmarks come from companies that sell survey tools (and so does WhiskrKit). Their customers are teams that already care about surveys, and each vendor defines “response” a bit differently. Use the numbers as a rough range, not as a target to hit to the decimal.

First, check how you calculate it

Two teams with the exact same survey can report very different response rates, just because they divide by different things.

  • Response rate is usually the number of survey views where someone answered at least one question, divided by the total number of views. That’s how Refiner defines it.
  • Completion rate is about who finished. Refiner counts views where every question was answered.
  • Alchemer is stricter for response: it only counts submitted surveys.

The denominator matters most. If you count every time the survey could have been shown (eligible users) instead of every time it was shown, your rate will look much lower. And if the same person sees your survey in every session, each view adds to the denominator while they’re only ever going to answer once.

The same survey gives a 2%, 4% or 6% response rate depending on what you divide by 60 people answered the survey. Divided by 3,000 eligible users that is 2%. Divided by 1,500 impressions it is 4%. Divided by 1,000 unique views it is 6%, the recommended way to count. Of those 1,000 unique views, 45 people answered every question, a 4.5% completion rate. Same survey, same 60 answers Three ways to count give three different response rates 60 ÷ 3,000 eligible users 2% 60 ÷ 1,500 impressions (incl. repeat views) 4% 60 ÷ 1,000 unique views COUNT THIS 6% Completion: 45 finished ÷ 1,000 unique views 4.5% Answered at least one question Answered every question What you divide by

So before comparing yourself to anyone: count unique views, decide whether “response” means one answer or a full submission, and keep that definition the same over time.

Why some surveys get 5% and others get 40%

1. Timing

When you ask matters more than almost anything else. Apple’s Human Interface Guidelines give advice for rating prompts that applies to any in-app survey: only ask after people have shown real engagement with your app, and don’t ask on first launch or during onboarding, because people haven’t formed an opinion yet.

In practice, the best moment is right after someone finishes something: they booked the ride, saved the document, completed the workout. Asking in the middle of a task means you’re competing with the reason they opened your app. We went deeper on this in How to Decide When to Show a Feedback Prompt in Your iOS App.

2. Length, and especially the first question

Common advice says shorter is always better. The data is a bit more interesting than that.

In Refiner’s 2022 data, single-question surveys clearly performed best. In their 2025 data, surveys with 4 to 5 questions had the highest response rate, with single-question surveys also doing well. Their explanation is that on mobile, tapping through a few quick answers fits into how people already use their phone.

SurveyMonkey’s research across a much larger set of general surveys adds two useful details:

  • Drop-off grows fastest with each added question in the first part of a survey, roughly up to 15 questions.
  • The number of words in the first question predicts lower completion, even after controlling for the total length of the survey. Hard question types also cost you: surveys with 10 matrix or rating-scale questions averaged 81% completion, against 88% for surveys with just one.

The takeaway: the first question decides whether people start at all. Make it short, concrete and answerable with one tap. Save the open text field for later, when they’re already invested.

3. Placement and format

Refiner found that surveys inside mobile apps got a higher response rate (36.1%) than surveys inside web apps (26.5%). On mobile, a survey at the bottom of the screen averaged 34.8%. But 86% of the mobile surveys in their dataset used that placement, so it’s more “the default works” than “the bottom is proven best”. Full-screen surveys scored even higher, but with only 9 surveys in the sample, that’s not much to go on.

The bigger gap is between in-app and everything else. Apptentive’s 2017 benchmark report compared its 13% in-app average with typical mobile survey response rates of 1% to 3%. A survey that feels like part of the app gets answered. A link to an external form mostly doesn’t.

That also means the look of the survey matters. Something that looks like a web page dropped into your app, with the wrong fonts, keyboard behaviour or dark mode, feels like an interruption from someone else.

4. Targeting: ask the right people, not everyone

Showing the same survey to every user is the fastest way to a low response rate. If you want to know what people think of your new export feature, ask the people who just used the export feature.

Jakob Nielsen of Nielsen Norman Group suggests a related trick for when you have many questions: split them across users. Instead of one survey with ten questions, make five short surveys and randomly give each user one of them. Each person answers less, and you still get all your answers.

Also set a frequency limit. If someone dismissed a survey yesterday, showing it again today mostly teaches them to dismiss faster.

5. What you ask about

In Refiner’s 2025 data, CSAT surveys (“How satisfied were you with…?”) got a higher response rate than NPS surveys (“How likely are you to recommend…?”): 26.3% against 21.7%. A plausible reason is that CSAT usually asks about something that just happened, while NPS asks people to judge your whole product. Questions about a concrete moment are easier to answer.

6. Letting people choose

Alchemer reported that surveys people could opt into, linked from an in-app message instead of shown directly, averaged a 50% response rate. Fewer people see an opt-in survey, but the ones who do actually want to answer.

Don’t chase the number blindly

Here’s the part most benchmark posts skip: a higher response rate is not automatically better data.

Nielsen argues that low response rates can mislead you, because the people who do answer tend to be your most committed users, not your average ones. That’s true. But the research on survey methods adds an important nuance. In a well-known meta-analysis of 59 studies, survey researchers Robert Groves and Emilia Peytcheva found only a weak link between a survey’s response rate and how biased its results were. A survey with 15% response can be more representative than one with 40%, depending on who answers.

So instead of only asking “how many responded?”, also ask “do the people who responded look like my users?” A simple check:

  • Compare respondents with everyone who saw the survey, by app version, platform, how long they’ve been a user and how active they are.
  • If your respondents are mostly long-time power users, your scores will look better than reality.
  • If they’re mostly people who just hit a bug, your scores will look worse.

A smaller group that represents your users well beats a big group that doesn’t.

What to fix first

If your response rate is low, work through this list in order. Each step is cheaper than the next.

  1. Check your math. Count unique views, not impressions or eligible users.
  2. Move the trigger. Show the survey right after a completed task, never at launch or during onboarding.
  3. Shorten the first question. One line, one tap.
  4. Narrow the audience. Ask people who actually used what you’re asking about.
  5. Add a frequency limit. Don’t show the same survey again to someone who just dismissed it.
  6. Make it look native. It should feel like part of your app, not a web form.
  7. Then test length. Try a single question against a short 3 to 5 question version.
  8. Check who answered. Compare respondents with viewers before you trust the result.

How WhiskrKit fits in

WhiskrKit shows surveys as native UI in iOS and Android apps, and through wrappers for React Native and Flutter, so a survey looks and behaves like the rest of your app instead of an embedded web page. You decide in your own code when a survey appears, so you can tie it to the moments that work: after a booking, after a finished task, after someone used a new feature.

If you want to try it in your app, get in touch.

FAQ

What is a good response rate for an in-app survey?

In mobile apps, published benchmarks range from about 13% to 36%. Above 20% is solid for most apps. Below 10% usually points to a timing, targeting or placement problem.

What’s the difference between response rate and completion rate?

Response rate measures how many people who saw the survey answered at least one question. Completion rate measures how many answered every question. A long survey can have a decent response rate and a poor completion rate.

How many questions should an in-app survey have?

Either one question, or a short set of 4 to 5 quick questions. Recent benchmark data shows both work well. Whatever the length, keep the first question short and easy.

Do in-app surveys get more responses than email surveys?

Generally yes. Benchmarks consistently show surveys inside an app getting several times more responses than surveys sent by email or opened in a mobile browser, mostly because they reach people while the experience is still fresh.

Sources