In today’s research world online surveys via email, conducted using customer lists or dedicated research panels, seem to be the default option. Speed, cost and convenience often mean they are the best route but sometimes alternative data collection methods can be superior. Below are some factors to consider before deciding if an online email questionnaire is right for your survey:
Can you get email addresses for your target audience?
It is sometimes the case that you can’t get email contact details for a representative sample of your audience. For example if a publisher wants to know the views of the readership of one of their magazines they would want to contact readers. They may have email addresses for subscribers but would probably not have any contact details for people who bought the magazine directly from retailers. To gain the views of their complete readership they would need to use another method, perhaps publishing a URL in the magazine for readers to access a web survey.
Have you got permission to send surveys to the email addresses you’ve got?
Laws vary from country to country but to be able to send surveys those “individuals contacted by e-mail for research have a reasonable expectation that they will receive a contact for research” (ESOMAR Guideline For Online Research). In practice this usually means that respondents must have opted in to future marketing contact when their email addresses were originally collected. If respondents haven’t given this permission for re-contact then sending them online surveys via email risks breaking the law. If this is the case you will need to consider other data collection methods.
Will an email survey generate a high enough response rate?
Response rates for email surveys can be extremely low, sometimes less than 1%. Good survey design can improve response but the proportion of those sent the survey who complete it may still be minimal. The lower the response rate the less likely your survey is to be representative of your target audience. Or, to put it another way, the more likely those completing the survey are to be odd in some way. If it is key for your achieved sample to be representative of your target audience then methods that potentially yield a higher response rate, like telephone research, should be considered.
Is my sample geographically clustered?
Does the audience you want to interview live within a relatively small area or visit the same place? For example, if you want to interview residents of a town it may be easier to obtain a sample of people through appropriate telephone directories and ‘phone them rather than try to get a suitable database of email addresses. Similarly, if you want to interview visitors to a shopping centre then the best approach may be to interview people face-to-face when they are there.
Does my target audience respond to email surveys?
Some groups don’t ideally suit email surveys. For example, Internet penetration amongst the elderly is lower than for the rest of the population so the risk of missing out on certain types of people within this group is high. Although virtually universally online in developed countries, teenagers are hard to engage via email surveys, often finding them boring and irrelevant.
Is my survey too long and complicated to be online?
Research suggests that online surveys lasting more than 20 minutes (about 50 questions) lead to drop out and lower quality response. If your survey is longer than 20 minutes you should first consider if you can streamline the survey to make it shorter or split it into 2 different surveys. If this isn’t possible then you should consider alternative approaches that facilitate longer questionnaires. For example, in-home face-to-face surveys with good interviewers can last up to an hour and still generate a good quality of response. However, they are considerably more expensive than online alternatives.