How many times in the last year have you participated in any kind of survey?
More than 10, from 5 to 10, from 1 to 5, not a single time.
Survey is one of the most popular forms of quantitative research. Why? It's cheap, it scales, it could be distributed through different mediums and it delivers comparable results.
Similarly to other research methods you need to start from the problem statement.
What do you want to find out? What hypothesis do you want to validate?
If you've used a qualitative research before - you already have a hypothesis formulated and now you want to validate it with more people.
Next step would be to identify appropriate respondents. Your sample size needs to be large enough to obtain comparable and trustworthy results. Remember that you need to have or receive permission to reach for these people. Nobody would appreciate survey from unknown company or individual.
Now it's time to design the survey. Here is a great crib sheet on how to design survey questions:
From here.
From here.
Learn more about surveys:
More than 10, from 5 to 10, from 1 to 5, not a single time.
Survey is one of the most popular forms of quantitative research. Why? It's cheap, it scales, it could be distributed through different mediums and it delivers comparable results.
How would you approach survey method?
Similarly to other research methods you need to start from the problem statement.
What do you want to find out? What hypothesis do you want to validate?
If you've used a qualitative research before - you already have a hypothesis formulated and now you want to validate it with more people.
Next step would be to identify appropriate respondents. Your sample size needs to be large enough to obtain comparable and trustworthy results. Remember that you need to have or receive permission to reach for these people. Nobody would appreciate survey from unknown company or individual.
Now it's time to design the survey. Here is a great crib sheet on how to design survey questions:
From here.
.jpg?t=1422649435825)
From here.
Basic rules of survey content:
- Short and easy to complete (max. 3 min to complete)
- Easy to understand questions
- Comparable answers (closed questions, scales)
- Reasonably short
- Clear
- Open-ended questions only if you can't live without them (max 1)
"Pay" for the effort or call to take an action
Filling out a survey takes effort, Be it 1 minute or 10 minutes - your respondents would always find a better way how to spend this time. Instead, they've chosen to spend time on your survey. To appreciate their effort - offer something in exchange. It could be a discount for your service, a report that not in a public access or any other thing that could be interested in your respondents. If you can't offer anything like this - just provide your respondents with a nice, big "Thank you" and a C2A on what to do next: read this blog article, go to our homepage, login to your account...etc.Test your survey and then - test again
After you've done with designing questions - show your survey to a colleague or any other person. Ask him or her to read out questions and tell you how they understand them. You need to make sure your questions will be understood correctly by respondents. When you're sure about the language of your questions - test survey again with sample answers. Check if answers you're going to receive will contribute to the solution of a problem (see point 1).Write intro
If you send a link to a survey in an email or a landing page - write few lines explaining what's the purpose of the survey. Give potential respondents context and if possible - incentive to take a survey. You can use your survey as the opt-in point to recruit respondents for other types of research.Test delivery
Send out the survey to few test addresses to check if and how it will arrive. Often surveys land in "Spam" folder. Also, check how your survey invitation email looks in the most popular email clients and services.Send it out
When you're happy with your survey you can send it out to the world. Most online survey tools enable you to monitor the delivery rate of your survey, response rate, bounce rate and so on. Remember not to change survey before it's complete otherwise data you'll collect will not be comparable.Analyze results
When results are gathered you'll need to analyze them. Most survey tools would do the work for you. It will present you answers and % of respondents who chosen them. But that's you who would need to interpret the results and answer your initial problem statement. Infographic is a nice way to present survey results.Use the knowledge
You should never make surveys for the sake of doing them. Every survey results should lead you to a set of concrete actions. You've learned that users do not rate your customer support much - fix it. Your respondents are happy with the UI of your website - cheer up your designers. Customers cannot find product documentation - follow up with them. Use the power of knowledge that you gathered with the survey to the good of your product and customers.Mind the medium
Think about how you will deliver your survey to respondents, how would they interact with it? Most surveys nowadays send via email. Respondent is usually asked to click on a link in the email to start the survey. What would happen if respondent will open the survey on his mobile? Tablet? Could your survey adapt to deliver the best possible experience to respondents?Learn more about surveys: