First response (rate): How does the first question in a survey affect its response rate?

James Harding
Head of Product
First response (rate): How does the first question in a survey affect its response rate?

How does each question type help or hinder the response rates of your survey? We crunched the data... so you don't have to!

Contents of this article

After you’e decided what you want your survey to be about, how you’re going to feed it into your broader strategy and what you’ll do with the data you get back, you’re going to need to write some questions (assuming you're not using one of our expert written templates, obvs.). 

Writing questions can be tough. 

Not only do you have to decide what you need to find out, but you also need some idea of the specific type of question you want. 

On doopoll, we give you the option of 5 question types at moment: 

  1. Yes/No
  2. Multiple choice
  3. Single slider (akin to the Likert scale)
  4. Multi slider (akin to the Matrix question type)
  5. Open text questions

A key metric for us – because we consistently see people switch to doopoll from other products for this reason – is maintaining a high response rate for surveys and polls created by our customers.

Other platforms promise 20-30%, but we regularly get our users 60%+ response rates. 

We’re super driven to ensure that you get higher response rates. Because when you get more responses, your data becomes more meaningful. 

So here’s a question: how does the first question in a survey affect response rates to the second question? 

To answer it, we pulled a sample of more than 3.5K surveys (anonymised) that were created by our users and removed surveys when they didn’t have more than 2 responses overall (usually this is people who are testing what doopoll does before creating a second survey to send out). 

What’s left is a sample of 3500 surveys analysed by their first question type and the drop off in response between Q1 and Q2. 

What the data from 3500+ surveys shows about question types

Statistically, the worst type of survey question you can ask first is a question offering an open text box. 

What follows is a list of our 5 question types. In this list, we have added the average response rate drop off between question 1 and question 2 in any given survey when the respective question type is chosen. 

Or, more simply, the percentage drop off each question causes when it is the first in a survey. 

  1. Yes/No – 3.0%
  2. Multiple choice – 4.19%
  3. Single slider (akin to the Likert scale) – 3.2%
  4. Multi slider (akin to the Matrix question type) – 6.7%
  5. Open text questions – 9.6%

Or, if your brain needs a helping hand putting that in order (from best to worst):

  1. Yes/No – 3.0%
  2. Single slider (akin to the Likert scale) – 3.2%
  3. Multiple choice – 4.19%
  4. Multi slider (akin to the Matrix question type) – 6.7%
  5. Open text questions – 9.6%

Dropoff rates by question type

What does this mean for your next survey?

Response rate is only one factor in deciding whether your survey is a success. Ultimately, successful survey creators know that it is the quality of the response that is most important. 

That said, a component of quality is the amount of data that you’re getting. Because if it is more representative of the size of your audience, there’s a strong possibility that your data is more representative as a whole.

And we know that representative data allows you to make choices and implement changes with a good deal of authority. 

In the interests of achieving that certainty, you should choose a question type with a high response rate for your first question, and factor in how well it performs against the second question’s response rate too. 

Use question types (yes/no or single slider) that don’t require much cognitive load (brain power) to answer at first. 

This has two positive impacts: 

  1. You get people into the survey-answerin’ mood quickly and they’re more likely to complete surveys
  2. You get clear data points right away. This helps you segment data more easily. 

Does this mean I shouldn’t use Open Text? 

Nope. You can feel free to use it. You may see a 9.6% drop off on question two though. 

A better way to think about this is: Imagine striking up a conversation with a stranger. 

You’re probably going to do best when you ask them ‘What time is it?’ or ‘Good day so far?’ rather than ‘What are your opinions on global affairs?’ as an opener. 

People need some time to warm up and ease into a conversation. 

A good practice all around is to think of a survey like a conversation. 

If you keep this in mind when you’re designing your next survey. You’re always going to do pretty well. 

Easy to use survey builder.
Create as many surveys and questions as you like.
3-4x more responses.
Responses appear in real time.
Try it free

Your next read