SurveyMonkey and Google Forms have put collecting feedback into the hands of anyone who has a device and 5 minutes. Unfortunately, just because you know you to login to Google Forms doesn’t mean you will collect accurate feedback from the respondents. Many jobs require employees to gather feedback from customers, stakeholders, colleagues, clients, etc. but most don’t have a basic understanding of how to design a survey that will help you create an unbiased and accurate questionnaire.
For arts marketers, it can be tricky to find the best time to engage a customer in a survey. I haven’t had a lot of luck after a performance trying to get people to fill in a survey when they are trying to get home, and I don’t want to interrupt their experience at a gallery with a clipboard or IPad. I recommend collecting survey data from arts patrons in the following ways:
- Ask them when there is downtime, such as prior to the doors to a hall open, when they are waiting in line for a drink at the bar or at intermission.
- Send a post-performance survey immediately following an evening performance or first thing the next morning. When the experience is still fresh in the patron’s mind is the ideal time to ask someone a few questions.
- Offer a deadline and an incentive. Keeping the window of time short to respond to the survey, and offering tickets to an upcoming arts event or a draw entry for a good prize will increase participation.
- Don’t over-survey. I know one arts organization that sends me a survey after every single show that I attend. The volume ends up irritating me and I don’t fill them in any more. I didn’t see a lot of change in their operations and I wondered if they were actually analyzing the survey data that was being submitted. Pick a window of time and do the survey for your sample and then take a break to avoid over-saturation.
Once you know who you are surveying and when you will offer them the survey, the real work begins. You now have to create a survey that is free of bias, effective, short and that will ultimately answer the question(s) you need answered.
5 tips for creating a great survey
- Tell them why. Make sure you have an introductory paragraph that explains why the hell you are asking them to take 5 minutes out of their day to do this thing, and also what you are going to do with the information. If you are offering an incentive – popular options are draws for gift cards, etc – add that in this section as well.
- Only ask what you need to know. Filling in a survey can be pretty boring and unless there is a great incentive, you will have many respondents abandoning the survey before they finish it. For example, if you are looking at your arts programming for next season, and want to run by some ideas by your subscribers, do that. But don’t ask them what shows they attended last season (you probably already have that data in your ticketing system), or their household income, or the car they own. The more relevant the question to the purpose of the survey (see point #1), the more they will want to respond to the question.
- Avoid double barrelled questions. One of the most common mistakes that I observe is asking people two questions in one and then expecting them to be able to answer. For example, here is an question from a SurveyMonkey blog post: “How useful will this textbook be for students and young professionals in the field?” What if the textbook is only useful for students, but not young professionals? Make sure you don’t force someone to answer two questions at once.
- Use a balanced scale for ratings. Recently I received a survey that only offered me three options in response to a significant number of questions – Excellent, good, not good. What if I am not sure of my answer or perhaps I think something is terrible? This doesn’t give me a wide enough range of options to choose from, so make sure to use a balanced rating scale when asking people to rate satisfaction. I recommend:Strongly agree / agree / don’t know / disagree / strongly disagreeThis offers two positive options, two negative options and one neutral option. The most widely used rating scale is the Likert scale, and you can read more about that here.
- Test, test, test. Make sure you send the survey to a sample of a few people to test the flow of the survey. Find out if it’s easy to answer, if the purpose is clear, if it’s not too long, and if the questions can be answered accurately and fairly. I use this testing tool as a checklist before I send out a survey.
Online survey tools are easy to use and very inexpensive to deploy, but can be costly if the data they provide isn’t accurate and you make changes to your work based on inaccurate survey results. Here are some more handy tools to help you make sure your survey works!
Best online survey tools for 2019
Harvard University Questionnaire Design tip sheet
11 Customer Satisfaction Survey Mistakes to Avoid