Online surveys are undoubtedly the most cost effective way to gather important data such as customer satisfaction and employee opinions. But how do you maximise your response rate? This is a problem incurred by most businesses undertaking online surveys and there is much literature available relating to the topic, all with varying facts and figures in relation to what response rate to expect, techniques for increasing responses, what length of survey should be used etc.
The simple truth is that there is no single guideline to follow when undertaking an online survey, as each survey has its own set of variables affecting its response rate. However, there are some common techniques used to enhance the likelihood that a potential survey respondent will complete your questionnaire.
Having undertaken hundreds of online surveys of customers, employees, businesses, membership groups and event guests, I began recording survey response rates and the types of key variables likely to affect them. The variables recorded were:
- Whether survey invitations were personalised (i.e. Dear John Smith v Dear customer)
- Length of survey (categorised into completion times of 1-4 minutes, 5-9 minutes & 10+ minutes)
- If a reward/prize was offered
- Whether or not the survey was sent to people with a known interest in the survey subject (i.e. job related, previous customer, linked to social interests etc.)
The figures shown below were derived from the data of 143 online surveys, which were completed by a total of 49,335 respondents. As a general process, potential respondents were contacted twice via email to encourage participation in the online surveys.
Please note that the majority of respondents had submitted their details to various databases or were customers of my clients; therefore, it is likely that they were aware of the brand/company prior to receiving the online survey invitation. People sending email invites to random sub-sets of contacts should not expect to achieve response rates as high as those listed below.
Taking into account the 4 variables and their options, there were 24 possible combination’s – this will make sense when looking at the dot-points below. The combinations with the 5 highest and 5 lowest response rates have been listed.
5 highest response rates
- Personalised email & 1-4 minutes to complete & reward offered & respondents had a known interest in the survey subject = 45.3{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Personalised email & 5-9 minutes to complete & reward offered & respondents had a known interest in the survey subject = 40.1{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Non-personalised email & 1-4 minutes to complete & reward offered & respondents had a known interest in the survey subject = 39.8{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Personalised email & 1-4 minutes to complete & no reward offered & respondents had a known interest in the survey subject = 39.2{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Personalised email & 5-9 minutes to complete & no reward offered & respondents had a known interest in the survey subject = 38.4{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
5 lowest response rates
- Non-personalised email & 10+ minutes to complete & no reward offered & respondents did not have a known interest in the survey subject = 8.2{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Non-personalised email & 10+ minutes to complete & reward offered & respondents did not have a known interest in the survey subject = 15.5{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Personalised email & 10+ minutes to complete & no reward offered & respondents did not have a known interest in the survey subject = 16.2{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Non-personalised email & 5-9 minutes to complete & no reward offered & respondents did not have a known interest in the survey subject = 19.4{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
- Non-personalised email & 10+ minutes to complete & no reward offered & respondents had a known interest in the survey subject = 19.5{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} average response rate
So what works best?
The above findings show some clear correlations between response rates and the variables affecting them. As a general rule, it appears that the following actions should ensure that you gain the highest possible response rate when undertaking an online survey:
- Personalise your email invitations. People are more likely to respond to emails that address them by their name rather than ‘Dear Sir/Madam’, Dear Colleague’ or ‘Dear Customer’. If you have the names of potential survey respondents, use them! This is easily done using simple software such as Microsoft Word in a mail-merge format sent via your Outlook account. It links directly to your database, removing the burden of typing every single name into each email invitation.
- Keep the survey as short as possible. In most cases, the shorter the survey – the higher the response rate. Obviously some online surveys need to be in-depth and cannot avoid being 10+ minutes in length; however, try to avoid asking questions that only gather irrelevant information. This is a common crime amongst survey developers when preparing questions for an online survey.
- Offer a reward. Everybody loves getting something for free, so try and offer a reward when sending out online survey invitations. Some experts such as Kurt Knapton, Executive Vice President of e-Rewards Market Research, suggest that you are more likely to gain a higher response rate if you offer every respondent a small reward rather than a chance of winning a large sweepstakes prize. His study found that offering 4,000 people $2 each to complete a survey achieved a 19.3{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} response rate. Offering the same volume of people with an entry into a $2,500 sweepstakes draw only managed a 12.2{43188a7dd839b6435400250daa1cfd1f7fa6a9f2f74b5d47d7c17eef7596ad2a} response rate.
- Find contacts that are likely to have an interest in the survey subject. This is no scientific discovery, but response rates are always going to be higher if the potential respondent has some interest in the survey subject. Job satisfaction surveys always gain high response rates because employees believe that their answers will influence beneficial changes. Likewise, repeat purchasers of a particular product/service will undoubtedly have more interest in a related online survey than someone who has only used it once or never at all.