Online Surveys: How To Maximize Your Response Rate

Online surveys are undoubtedly the most cost-effective way to collect important data such as customer satisfaction and employee feedback. But how do you maximize your response rate? This is a problem faced by most online survey companies and there is a lot of literature available related to the subject, all with varying facts and figures regarding expected response rate, techniques to increase responses, the length of the survey to use, etc. .

The simple truth is that there is no single guideline to follow when taking an online survey, as each survey has its own set of variables that affect your response rate. However, there are some common techniques used to increase the likelihood that a potential respondent will complete your questionnaire.

After conducting hundreds of online surveys of customers, employees, businesses, member groups, and event guests, I began tracking survey response rates and the types of key variables that were likely to affect them. The registered variables were:

1. Whether the survey invitations were personalized (i.e. Dear John Smith vs. Dear Customer)
2. Survey duration (classified into 1-4 minute, 5-9 minute, and 10+ minute completion times)
3. If a reward/prize was offered
4. Whether or not the survey was sent to people with a known interest in the survey topic (i.e. job related, former client, linked to social interests, etc.)

The figures shown below were derived from data from 143 online surveys, which were completed by a total of 49,335 respondents. As a general process, potential respondents were contacted twice by email to encourage participation in the online surveys.

Note that most of the respondents had submitted their data to various databases or were clients of my clients; therefore, it is likely that they were aware of the brand/company prior to receiving the invitation to the online survey. People who send email invitations to random subsets 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 combinations; this will make sense when looking at the points below. The combinations with the 5 highest and 5 lowest response rates have been listed.

5 Highest Response Rates

1. Personalized email and 1-4 minutes to complete and reward offered and respondents had a known interest in the survey topic = 45.3% average response rate
2. Personalized email and 5-9 minutes to complete and reward offered and respondents had a known interest in the survey topic = 40.1% average response rate
3. Non-personalized email and 1-4 minutes to complete and reward offered and respondents had a known interest in the survey topic = 39.8% average response rate
4. Personalized email, 1-4 minutes to complete, no reward offered and respondents had a known interest in the survey topic = 39.2% average response rate
5. Personalized email, 5-9 minutes to complete, no reward offered, and respondents had a known interest in the survey topic = 38.4% average response rate

5 lowest response rates

1. Non-personalized email and more than 10 minutes to complete and no reward offered and respondents had no known interest in the survey topic = 8.2% average response rate
2. Non-personalized email and 10+ minutes to complete and reward offered and respondents had no known interest in survey topic = 15.5% average response rate
3. Personalized email, 10+ minutes to complete, no reward offered, and respondents had no known interest in the survey topic = 16.2% average response rate
4. Non-personalized email and 5-9 minutes to complete and no reward offered and respondents had no known interest in survey topic = 19.4% average response rate
5. Non-personalized email, 10+ minutes to complete, no reward offered, and respondents had a known interest in the survey topic = 19.5% average response rate

So what works best?

The above findings show some clear correlations between response rates and the variables that affect them. As a general rule, it seems that the following actions should ensure that you get the highest possible response rate when taking an online survey:

1. Personalize your email invitations. People are more likely to respond to emails that address them by name rather than ‘Dear Sir/Madam’, Dear Colleague’ or ‘Dear Customer’. If you have the names of potential respondents, please use them! This is easily done using simple software like Microsoft Word in a mail merge format sent through your Outlook account. It links directly to your database, removing the burden of writing each name on each email invitation.
2. 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 comprehensive and cannot help but take longer than 10 minutes; however, try to avoid asking questions that only collect irrelevant information. This is a common offense among survey developers when preparing questions for an online survey.
3. Offer a reward. Everyone loves getting something for free, so try offering a reward when you send out online survey invites. Some experts, such as Kurt Knapton, executive vice president of market research for e-Rewards, suggest that you’re more likely to get a higher response rate if you offer each respondent a small reward rather than the chance to win a large prize of lottery. His study found that offering 4,000 people $2 each to complete a survey achieved a 19.3% response rate. Offering the same volume of people with an entry into a $2,500 raffle only achieved a 12.2% response rate.
4. Find contacts who are likely to be interested in the survey topic. This is not a scientific discovery, but response rates will always be higher if the potential respondent has some interest in the survey topic. Job satisfaction surveys always get high response rates because employees believe their answers will influence beneficial changes. Similarly, regular buyers 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.