It’s no surprise that my #NAMPC coaching sessions about creating surveys are always filled. When it comes to surveying, you, like most people, probably have the most trouble with simply getting started. These five tips should help you, if you do them in order.

1.Set your objectives.

What’s the real purpose of your survey? Your first step should always be getting clear on what your results will be used for and who will use them. Questions to ask yourself:

Will results be used internally or externally?

What decisions can be made based on your survey results?

Are you testing a hypothesis?

Are you reporting on the impact of a program for a grant?

Setting your objectives by answering these questions will start you down the right path.

2.Define the type of report you want.

What should your report include? It may seem illogical to start thinking about your report before you even have survey questions, but it’s not. Asking yourself these questions before designing your survey will keep you on track.

Do you want to report percentages of things?

Do you want to rate or rank things?

Do you need to understand influences or correlations between things?

What groups of people do you need to compare results across?

How complex does your analysis need to be?

If you just need to report simple percentages of things, rate things, and rank things, even across groups such as subscribers and non-subscribers, that’s not too tough. It will be easy for you to pick your question types (step 5) and use a tool such as SurveyMonkey.com or QuickTapSurvey.com to automate your analysis. If you need to understand how one thing influences another or relationships between things (such as patrons age and programming preferences), then you will need to be extra careful in your question design and get assistance with analysis using statistical software (such as SPSS, or SAS). Consultants and students in graduate math programs can help with this.

3.Design your process.

How will this project flow? Again, before even thinking about your survey questions you need to map out your process.

From whom will you collect data?

What data will you collect?

How/where will you be most successful in collecting your data?

When will you administer your survey?

Mapping out the mechanics of your process will help you decide what kind of data you can get.

4.Decide what kind of data you want.

Did you know there are really only four kinds of data you need? If you think about your data needs (based on your objectives, reporting, and design process) in the following four categories, your survey questions will come easily.

Knowledge: These find out how much people know about or understand something. (E.g. is our organization, a) a commercial enterprise, b) a nonprofit organization, or c) a municipal facility?)

Opinions: These reveal what people feel or believe about something. They can ask about the past, present or future. (E.g. was this program a good value for the cost?)

Behaviors:  These ask about what people do or expect to do. (E.g. would you recommend this program to other people?)

Characteristics: These document the attributes of people or households taking your survey. They are typically demographic, geographic, status with your organization, or proxy for lifestyle insights. (E.g. Which store would you most prefer to shop for clothing for yourself? Nordstrom, Forever 21, the Gap, or TJ Maxx?)

There are many types of questions you can create in these four categories. Keeping you question types linked to your objectives will help you pick your question types.

5.Pick your question types.

There are so many question types, where do you start? Picking your question types becomes a lot easier once you’ve completed the first four steps. I promise! Here are a few popular question types you can choose from before you sit down to write your survey.

Close-ended questions: These are single choice or multiple choice lists of items such as yes/no, rating scales, or pick your favorites. (E.g. how strongly do you agree or disagree with the following statement?) These can be single questions or multiple questions structured in a matrix of rows and columns.)

Partially closed: These offer an “other” category after closed-ended questions. Use sparingly.

Paired: These are questions that help you compare things (E.g. Comparing classical symphony orchestra concerts to the following other types of concerts, which types of concerts do you generally prefer? Options: chamber music, pops, opera, world music, jazz, etc.)

Open-ended: These give people the opportunity to self-report and self-express. Try to limit your survey to one or two of these. They are great for capturing stories but difficult to summarize in reports.

Most people start their survey process by trying to write questions. That’s where you’ll get stuck. If you follow these first five steps, writing the questions becomes much easier!

Next up: Tips for writing your survey questions, question pitfalls, administering your survey, analyzing your data, and reporting your results – so stay tuned!

What is your biggest challenge in starting a survey project? We’d love to hear from you. And, if you would like more personal help with starting surveys, schedule a free coaching session with me at #NAMPC or come on my dine-around and we’ll sort it out together!

See you in November!

Surale Phillips is providing one-to-one coaching on the topic  Designing Audience Surveys with Purpose during the National Arts Marketing Project Conference 2015. REGISTER before this Friday, 10/23!