Optimising online survey invites

The way you introduce surveys to respondents can have a significant impact on their success. It doesn't matter how well constructed your questionnaire is if you can’t get people to complete it.

Survey invites should be designed to get as many people as possible completing your survey. The better the response the quicker you can get the interviews you need. Low response rates may delay your project and have cost implications if you are buying in sample or using a panel.

However a good level of response to your survey is only worthwhile if you’re interviewing the right mix of people. Your sample can be skewed if the invite is more appealing to niche groups within your wider target audience. You also don’t want to make it easy for less scrupulous respondents to qualify for the survey and any associated incentive by providing false answers.

Here is our list of dos and don’ts for online survey invites:

DOs:

  • Do personalise the introduction where possible. Start “Dear Fred” rather than “Dear customer”.
  • Do tell respondents how you obtained their contact details (unless this information potentially impacts on how respondents answer a survey, in which case details should be given at an appropriate point in the interview).
  • Do make the survey sound interesting and relevant to respondents.
  • Do explain why you’re doing the survey and what the results will be used for. Tell respondents why their opinions are valuable. Respondents will be more motivated to complete a survey if they understand its purpose.
  • Do add the signature and job title of someone prominent in the client organisation to the invite. The more high profile the name the better. People are more likely to think the survey is important if it is from the CEO rather than, for example, a Marketing Assistant.
  • Do provide an email address and/or telephone number through which respondents can contact you with any problems or queries. Just providing contact details makes the invite appear more trustworthy and those who make contact will probably complete the survey when they otherwise might not have done.
  • Do conform to industry guidelines on how to present surveys to respondents.The Market Research Society Guidelines for Questionnaire Design can be accessed here while the ESOMAR guidelines “Using Online Identification and Tracking Technologies in Research” are available here.

DON’Ts:

  • Don't tell respondents the survey will be shorter than it actually is to get them to click on the survey link. Studies have shown that interview length has little impact on response rates. If respondents want to complete the survey, knowing it will take 15 or 20 minutes rather than 5 minutes does not put them off. However, lying to respondents by saying it will only take 5 minutes may cause them to drop out of your survey and deter them from taking part in future research.
  • Don't be specific about the type of people you want to interview in your invite. This will give cheating respondents the information they need to provide false answers to qualify for your survey.
  • Don't use language that particularly appeals to only part of your target audience. For example, an invite for a survey about cars could say something like “Calling all petrol heads, tell us about what you drive”. This sort of language is more likely to attract those who are really interested in cars, people more likely to drive a lot and have high performance cars. This might be fine if you are looking for a niche audience but if you are looking to interview all drivers your sample will probably not be representative of this group. A better wording would be something like “We’re interested in how people get from A to B and the different types of transport they use”.

Related posts:

To pay or not to pay: using "incentives" in survey research

Submit comment

Recent Posts

Twitter