Surveys are great ways to collect data about pretty much anything to do with your website for example, how people use your website, what they think about your services, how they rate your products. However, writing a good survey with results you can analyse and data you can actually use is not as simple as many people think.
Below is some advice and a few tips on writing a survey for your website…
No one wants to be faced with dozens of questions especially online where attention spans continue to fall. To ensure as many questions are answered as possible and reduce the drop out rate aim for 10 questions or less. If you decide to split the questions in to multiple pages don’t create so many stages it feels like they are climbing a mountain and always let them know where they are in the process e.g. step 2 of 5.
Give them a reason to take part
Provide a strong incentive for the participant to complete the survey and give you their contact details. For example, if you are running a survey to benchmark your industry you could give them access to the results or if it is a survey about your service a competition is always popular e.g. Win a laptop.
Structuring the questions
How you structure your questions will make the difference between having data you can apply and having a large spreadsheet sitting untouched on your computer. Some tips include…
1. Don’t weigh the answers in favour of a certain answer e.g. Excellent, Very good, Good, Poor (3 positive and only 1 negative)
2. Don’t put “don’t know” or “NA” as an option in middle (e.g. 3 of 6), it will make statistical analysis a nightmare. Put it as a separate option at the end
3. Avoid leading questions e.g. “Do you agree that…?”
4. Keep the question specific e.g. don’t use words such as “Occasionally”, “Often”, “Regularly” etc. What they mean differs from person to person
5. What does average mean? Don’t use it for the sake of it
6. Don’t fall in to trap of asking two questions in one sentence. “Did you buy our red balloon and what did you think of it?”
Open and closed questions
Open questions allow the respondent to answer a question with no limitations on what they can say e.g. “What do you like about your current supplier?” A closed question has a limited number of responses to choose from e.g. “How would you rate our customer service?” Very Good, Good etc.
In theory open questions provide you with more qualitative data but they can be difficult to analyse. If you have thousands of responses you will have to go through each one grouping the answers to be able to spot any common themes. Closed questions make it very easy to analyse the data but you could miss out on some valuable feedback.
Piloting the survey
Good practise dictates you should run a pilot survey to test the questions amongst a limited sample before launching it to the general population. Doing this will highlight any problems with the questions and any in data collection and analysis.
Next Steps
If you are interested in online surveys or think your clients maybe, call us and we can discuss options and costs.
