There used to be only three ways to do surveys: in person, over the phone, and on paper.
Ah, such simple days!
Now we've got web surveys and e-mail surveys and IVR surveys. And we've got hybrids like e-mail notifications with a link to take a web survey. So what's the best method?
The answer, of course, is that "it depends." There are two main considerations: who might be excluded from the survey (which leads to bias), and what response rate you might get.
These days nearly everybody has a telephone. Internet and e-mail isn't universal yet, but the majority of people now have access, at least in the United States and many other developed countries.
Lack of phone, web, or e-mail access isn't the only reason people may be excluded from a survey, though. For example:
- E-mail: Surveys embedded in an e-mail might not display properly in all e-mail software. Some e-mail readers will strip off embedded HTML forms, and even embedded links. This is becoming more and more common as a security measure. E-mails with embedded surveys or links to surveys also often get marked as spam, and never make it to the intended recipient.
- Web: Web-based surveys are more likely to work than e-mail surveys, but you still need to be careful to make sure the questionnaire works on all common platforms. The bigger issue is how to get the survey in front of the participant. Depending on how you display the link to the survey, it could get lost in the clutter of a web page, and e-mail notification often doesn't get through.
- Phone: Phone surveys may seem like they pose few technical obstacles, but automation tends to gum up the works. Surveying customers at the end of a customer service call (the "stay on the line" survey) excludes anyone who gets upset or frustrated and hangs up early--and also tends to be easy for customer service representatives to manipulate. Many people now have "privacy guard" services on their phones which block automated calls (as well as the predictive dialers often used to increase the efficiency of live interviewers), and may simply refuse to answer if they don't recognize the caller ID.
Response rate can vary dramatically from survey to survey, depending on (among other things):
- Timeliness: Surveys administered immediately after some customer interaction have a higher response than surveys offered days or weeks later.
- Reach: Many surveys never get to the intended participant, because they don't answer the phone, the e-mail gets marked as spam, etc.
- Impact: Customers are more likely to answer a survey when they believe that the company will pay attention to the results.
- Convenience: Overly long surveys, or surveys offered in a time or manner which isn't convenient to the customer, will often be refused.
With all these factors, it's no wonder that the best way to administer a survey will depend on the circumstances. Here are some rules of thumb:
- Respect your customers, and always remember that you're asking them to do you a favor. Keep the survey as short as possible.
- Survey immediately after a customer interaction (when it makes sense for the survey goals).
- Survey in the same medium as the customer interaction. For example, if the survey was triggered by a phone call, you should survey over the phone. If the survey was triggered by a web site visit, use a web survey.
- If the survey isn't triggered by a customer interaction, consider a paper-mail survey. This tends to be the most convenient for customers, and can have a very high response rate.
- Don't survey at the end of a phone call (the end-of-call, or "stay on the line" survey). It excludes the customers you most want to hear from: the upset, frustrated, and dissatisfied ones.
- For phone surveys, use a live interviewer. Automated phone surveys often leave people feeling cold, and get blocked by common anti-telemarketing services.
- Tell participants why you're surveying and what the data will be used for. Convince them that answering the survey makes a difference.