Editor’s note: Dan Rangel is director of research solutions at research firm Survox, Chicago, and president of the Great Lakes Chapter of Marketing Research Association. This is an edited version of a post that originally appeared here under the title, “5 easy best practices to optimizing your IVR survey.”
Leaders in market research, customer CX and customer insights have found automated phone surveys (IVR) to be an essential tool in reaching out to customers to gain the crucial insights needed to develop and deliver incredible products and exceptional service.
That’s the good news.
However, as a career market researcher, I can say with some authority that many of these proponents of IVR surveys are not privy to some simple performance-enhancement techniques.
Stay focused and limit the range of multiple choices associated with each question: Overwhelming your respondents with too many choices per question is a surefire strategy to squash participation rates. Based on our field experience, we recommend you limit the number choices respondents need to make to no more than five. If you need to explore more choices, split the question into a 2a and 2b scenario, each limited to no more than five choices.
Another best practice tip to keep in mind when crafting IVR surveys with multiple choice lists: Always include the pound sign as a choice to repeat the selections. Respondents will thank you with a higher participation rate!
Present your questions in the most effective order: Your survey should flow logically and transition smoothly from topic to topic. Jumping back and forth from demographic-related questions to questions that address the experience of the product or service could confuse the respondent and contribute to lower completion rates. It’s important to keep in mind that surveying your customers is an integral part of the customer journey and that done well, surveys can create a positive and memorable branding experience.
When quota matters, include screeners: When you need a representative sample – respondents that meet specific demographic requirements such as age or gender, or who have previously experienced the product or service – you need to include screening questions. Screeners come at the start of a survey as means to determine whether respondents have the needed characteristics to make them eligible to continue with the survey. Properly deployed, screeners help ensure that the survey sample includes only those respondents that can provide helpful feedback based on study’s research goals.
Incorporate intuitive, respondent-friendly rating systems: Success in achieving higher participating rates is in the details. Setting up a respondent-friendly rating system is a detail that we have found crucial to helping our clients optimize their IVR surveys. Respondent-friendly means you limit your scale to single-digit numbers; we highly recommend a range of 1 to 5, no more than 7. And if at all possible, avoid using two-digit numbers such as 10 or 12.
A respondent-friendly rating system is also intuitive. What this means is that the number the respondent presses corresponds to the rating. On a scale of 1 to 5, the respondent would press 1 for the lowest rating or 5 for the highest. Simple, but a detail that if overlooked, can potentially result in bad data and a less than optimum user experience
The optimum number of questions: While there is no hard and fast rule, our experience in the field has been that completion rates tend to drop as the survey gets longer. With so much competition for the respondent’s attention, this nugget of knowledge is hardly surprising. But I say there are no hard and fast rule because the context of the survey is a major factor.
On average though, it’s a good practice to time your survey and try to keep it under five minutes.