This article explains the concepts of response rate and drop-off, shows you where to find these numbers in Skyra, and provides practical tips on how to improve them. By analyzing response rate and drop-off data, you can make small adjustments that increase completion rates and provide more value from feedback.

Response rate

Response rate is the percentage of users who have seen the survey and actually answer at least one question. For example, if 100 people see a survey and 3 of them respond, the response rate is 3%.

What is a typical response rate?

Our customers usually have a response rate of between 1 and 5%. In some cases higher.

Example: Sparebanken Vest achieved a response rate of over 5% by placing a survey on a logged-out page after online banking. The high response rates are due to relevant timing and targeted questions.
Read customer story

Here you can see the response rate in Skyra

Survey

Go to Survey > Select survey > Overview. The response rate is shown in the row of key figures at the top.

In the Performance tab you will find the response rate highlighted in blue in the timeline. Move your mouse over to see numbers per time unit.

Drop-off

Drop-off is the proportion of users who started to respond but did not complete the survey. Drop-off can be measured from one question to the next or from start to finish of the survey.

If 100 people answered the first question and 70% answered the last, the drop-off is 30%.

What is the usual drop-off?

This is influenced by several factors and between customers, surveys and websites.

Example

It's not uncommon for 60-70% of respondents to answer question number 5. In other words, 3 out of 10 have dropped out along the way.

Here you can find figures for drop-off in Skyra

Drop-off shows up where the numbers are different. For example, fewer people have responded to question number 5 than to question 1.

Four places you can identify drop-offs are in Survey report, Survey Performance, Responses table and Task List.

Remember question logic
Follow-up questions have fewer answers because they are only shown to some, based on previous answers in the survey. The setting for this is Visilibily rule.

1. Survey Report

Go to Survey > Select Survey > Overview.

The survey report shows scorecards for questions in the survey. By seeing how many fewer answers the next questions have, you can calculate the drop-off from a previous question.

2. Survey Performance

Gå til Survey > Velg Survey > Performance.

How many percent have answered a question?

At the bottom of this page, you'll see how many responses each question received, and what percentage of those who started the survey answered each question. This is essentially the opposite of drop-off, but aimed at the basic question "How many answered the question?"

In the example above, we see that 81.8% have answered question 3. That means that 19.2 percent dropped off after a first question. Some questions have zero answers, but there are also questions that are not shown to everyone. They have a Visibility rule, which means that they are only shown to some, based on answers in the previous question.

3. Responses table

In the table, we see that some columns do not contain answers from the user. There are three possible reasons for this.

Question rule: Because the question was not shown to them, based on question logic in the survey. For example, question 4 was only shown to those who answered "no" in question 3.
Optional question: Free text questions can be optional. If the user clicked through without leaving a comment, it's not a drop-off, but rather that they didn't have any feedback to leave.
Drop-off: If a mandatory question is not answered and the rest of the questions in the survey are not answered, this indicates drop-off. In other words, the user did not complete the survey. Row three in the example above shows this.

4. Task List

The Task List shows numbers for two important questions in a top task survey.

What did you come to the website for? (Select top task)
1. The number of people who answered question 1 is counted as Responses
Did they do what you came for? (Yes/No)
1. The number of people who answered question 2 is counted as Completed or Failed, based on whether the answer was yes or no. The drop-off appears when there are fewer answers to question 2.

Drop-off appears when there are fewer answers to question 2. In the image above, we see that the number of Responses is 2133 and the sum of Completed and Failed is 1484. So the drop-off is 649 users, or 30% (649/2133=0.3).

Drop-off per top task

With View settings the columns in the Task List can be customized to show the number of Completed, Failed and Responses. With columns showing numbers for questions 1 and 2, it's easier to see what the drop-off is per top task.

Reasons for drop-off and tips for improvement

Minimization
The user has minimized the survey and forgets to open it again to respond.

Tip: Avoid automatically minimizing the survey and make sure there is contrast in the colors of the minimized survey and the website, so it is easy to see.

Leaving the website
User navigates to an area of the website where the survey does not follow (e.g. logged in service or new domain).

💡 Test that the survey actually follows suit in other areas/services on the website. Also check that the URL rules for the survey are correct.

The survey is too long
Surveys that are too long often cause people to give up halfway through.

💡 Limit the number of questions to what is absolutely necessary, max 7-8 questions in a pop-up survey on the website.
💡 Prioritize questions that provide the most value, as dropouts often happen early on.
In a top task survey, it's a good idea to start with questions about Top task and move on to questions about Solution rate. Then you get the answers you need to build the statistics in the Task List.

Questions are unclear or demanding
Input fields collect fewer responses than questions with options. It's not uncommon for respondents to continue without answering.

💡 If it is important that people respond, you can set an input card to require at least 3 letters before you can press on.

Relevance
If the questions are not perceived as relevant, respondents will lose interest.

💡 Avoid vague questions. Instead of asking "What do you think of our website?", ask "On a scale of 1-5, how easy was it to use the service?". Ask those who gave a low score "What can we improve?".
💡 Make sure the questions are logical and flow naturally

Timing
If the pop-up appears at the wrong time, for example immediately after the user arrives on the page, it can lead to a lower response rate.