How To Conduct Effective Online Surveys
 by: Nick Hill

Although they are inexpensive and product fast results, online surveys still need to be conducted properly to be really effective. The key steps for conducting effective online surveys are:

    * Identify your objective
    * Decide what information you need
    * Develop a questionnaire
    * Conduct the survey
    * Analyze the responses
    * Recommend a course of action

* Identify your objective

Online surveys are conducted to find out the characteristics, behaviors, opinions, and knowledge of a particular population. No matter what the subject matter, all surveys are conducted with an objective, to gather information relevant to a specific problem or situation. And the key to a successful survey is clearly identifying that objective upfront. To do that, ask yourself "what problem am I trying to solve?" The ideal objective has a narrow scope and can be clearly stated. The wider the scope, the more complex your survey and the less meaningful its results. Always remember, your survey is only as good as your understanding of your objective!

* Decide what information you need

Once you have defined your objective, identify the information you need to gather to achieve your objective. If the objective is clear and narrow, the information you need to gather will be apparent. Be specific and avoid the temptation to gather "extra" bits of information that are "nice to know" but not relevant to your objective. Once you have identified the information you need, you should explore the possibility of gathering it from existing data sources. With an abundance of information available from corporate databases and public information sources, you might gather the information without even conducting a survey!

* Develop a questionnaire

Developing the questionnaire is undoubtedly the most important part of conducting a survey. The quality of the questions will determine the quality of the results and the effectiveness of your survey. Get tips for developing an effective questionnaire.

* Conduct the survey

Once you have developed the questionnaire, you need to test it on a few people to determine if it is clearly written and can be understood by people representing your potential respondents. Testing the survey might take several iterations, but it is well worth the effort. Once you are satisfied that your questionnaire is clear and unambiguous, you are ready to conduct your survey.

If you are going to distribute your survey through email, make sure that your email message is clear and compelling. A boring message will not get the attention of your potential respondents and decrease your response rate. Additionally, you should also remember the following things:

    * Address emails to a single person whenever possible.
    * Always use the blind carbon copy (BCC) if you must send a single email message to a group of people.
    * Include a valid email address in the "From" field or the recipients will consider your message as spam.
    * Provide the URL that will take people directly to your survey, if you are inviting them to the website to take the survey.
    * Tell recipients how to contact you if they have a problem or concern.
    * Identify the source of purchased email addresses so recipients know who to contact if they want to be removed from the list.
    * Ignore the "flame" or "hate" messages you might receive in response to your email invitation. A small percentage of the recipients always take offense to something or the other.

If you are going to distribute your survey from your website, you need to consider how will get visitors to take the survey. If your site has just enough traffic, you can probably place a hypertext link on your home page through which visitors can reach your survey. If you have very heavy traffic, you can use a program on your web server to randomly deliver your survey to some of the visitors. However, if your website doesn't have enough traffic, you should consider sending invitations by email or other traditional methods.

Once you have sent out notifications, responses will begin to come in immediately. For online surveys that invite individual participants (email or postal mail), most people will respond within 72 hours of receiving your invitation. Surveys that rely on passive participation (clicking on website links or responding to indirect advertising) generally take much longer. In either case, you need to decide how long to keep your surve active based on your target audience and how many responses you need to collect. If you are surveying a known group of people (e.g. customers, employees, students, etc) you might need to send out reminders to remind some people to take the survey.

* Analyze the responses

You can begin to analyze the responses as soon as you receive them. Unlike traditional methods, online surveys make it possible to analyze responses without human intervention. Responses are received in electronic format and immediately added to a database. Once in the database, they can be manipulated with a variety of tools, including the survey software, statistical packages, spreadsheets, and presentation programs. The key to proper analysis of online surveys is no different than that for traditional methods. You must be able to filter the data and display results in a variety of formats for effective analysis. And always remember to relate the analysis back to your objective.

* Recommend a course of action

The last step in conducting effective online surveys is to recommend a course of action. With your objective in mind, you should be able to recommend specific actions as a result of your survey. Your recommendations can vary widely depending upon the nature of the problem you are trying to solve. If you cannot identify any actions as a result of your survey, then it is highly possible that your objective was too broad, your information needs were incorrect or incomplete, or your questionnaire was poorly written. In these cases, your recommended action should be to conduct further study.

© Web Based Survey Software (http://www.web-based-survey-software.org) 2004

About The Author

This article has been copyrighted and has been reprinted with the permission of Web Based Survey Software (http://www.web-based-survey-software.org). If you want to reprint this article, please send an email to [email protected] to inform us and we will grant you the permission. But please ensure that this message appears at the end of the article and proper credits are given.
[email protected]