Before you can design an effective survey, you'll need to define your goals. These include deciding what information you most want to gather, whom you are targeting, how long the survey will run, what will constitute a successful survey, and what you'll do with the information that you gather. Remember to include your existing customer data in the equation.
Define your primary and secondary goals. For example, you may develop a survey about your Web site. The primary goal might be to get feedback on what people like and dislike about it so that you can improve it. A secondary goal might be to involve potential customers in your site and to introduce them to your products. If your goal is to prove a certain hypothesis about customer behavior, you'll need to make sure to ask questions that can shed light on the relationship you want to examine. For example, if you think that your younger shoppers may have a preference for a certain color of product, you'll need to ask questions about both color preferences and age in order to be able to draw any conclusions about the relationship between these factors.
You may want to target a specific audience for your survey, such as people who have ordered from you once but have not returned to your site. You'll need to identify techniques for reaching only that audience. To send your survey to your target audience, you have to know their e-mail addresses and send each one the actual survey or a link to your on-site survey. If your survey is targeted to a specific audience, you must tailor the presentation style and approach to them.
If you are looking for ideas, visit the Gallup Organization site. Gallup surveys people on more than just political issues and has many surveys running at any given time. You can see current results and archives of previous polls.
You will also have to decide how long you will run the survey. This will depend on how many hits your site normally gets and how statistically accurate, valid, and reliable the survey results need to be. Most surveys can be run for a couple of weeks or a month. If you let them run too long, people lose interest in them.
Finally, you will have to decide how you will use the information that the survey generates. Who gets to see it? Will you post it and share it with your visitors? This may be a good idea if you are trying to generate interest in your site, but may not be if you are asking how much people would be willing to pay for your products. Keep in mind that if you do a survey, participants will probably want to know about the results and expect you to do something visible with the information they gave you. If you ask them to rate your Web site and suggest changes, you had better tell them what you found, both the good and the bad, and what changes you plan to make.
Designing a survey requires a lot of planning. But the more effort you put into this stage of development, the easier it will be to create a good survey.