Internet Surveys

Internet surveys have become a preferred method in certain situations. They are effective at getting information fast; the bulk of responses come in the first few days. Since data collection is electronic, all the information can be tabulated immediately and frequencies can often be tracked in real time.

In some cases, Internet surveys can be less expensive than other methods because the data collection and handling is automated. If sample must be procured, the total cost of Internet surveys is the same, if not higher, than telephone surveys.

Internet surveys are especially good in situations where the respondents are known to you and have an interest in the subject (and they have e-mail or Internet access), such as employee, membership, or customer surveys. They are the best way to survey hard to find respondents as sample firms have established “panels” of respondents whose characteristics are in a database and have agreed to participate in surveys. They are also excellent for surveying the technically savvy, Internet surveys can be constructed with complex skip patterns and question rotation as in computerized telephone surveys. Internet surveys are also unique in their capability to add visual images to text, making it possible to test advertising, logos and even video.

Pros	Cons
Data is tabulated automatically so results can be available days or even hours after a survey is sent.	Respondents must volunteer to participate; they must “opt in.”
Very specific market segments can be targeted.	Response rates are typically very low. Large numbers of respondents need to be invited to participate, and large amounts of sample need to be purchased.
Surveys can readily be conducted anywhere.	Responses to Internet surveys are representative of respondents with Internet access, but not necessarily non-Internet users. One can never be sure who the respondents are.
Complex surveys with multiple skips are possible. Advanced methods, sorts, conjoint, and a variety of other methods can be integrated into the questionnaire.	Statistical estimation of representativeness is impossible.
The survey is not intrusive; it can be completed at the leisure of respondents.	Many available e-mail samples consist of respondents that receive many surveys and may complete surveys (perhaps inaccurately) just for the incentive.
Visual elements, photographs and video can be integrated into or readily tested by the research	Response rates for business-to-business surveys can be especially low.
	Panel members can be overused and become professional respondents.