2200 Powell Street, Suite 1010, Emeryville, CA 94608 contactus@pinpointresearch.com 800.776.7265

The Value of Open Ends – Never Survey Without Them!

It’s amazing the number of IVR surveys that fail to capture true customer sentiment because they don’t include open-ended questions. The best way for call centers to improve their own performance and enhance customer loyalty is with unbiased feedback. The only way to get truly objective feedback is to encourage customers to respond in their own words, no holds barred.

The Basics of Bias

Let’s face it, all surveys have some bias. Automated IVR systems eliminate the human factor so they don’t introduce interviewer bias in the responses, but the nature of the questions themselves skews the results. Multiple choice questions still limit the response choices, so you only measure based on the limited parameters you create. Garbage in, garbage out. Bias often is introduced when creating the survey, especially if you are looking to validate a preconception or looking to generate a specific outcome. Too often surveys are structured to meet management expectations rather than getting unbiased feedback.

Let’s consider two of the basic biases typically introduced in multiple-choice surveys:

  1. Answer option bias – When presenting a list of potential responses, where the response falls in the list can influence the response. If the respondent is in a hurry, there may be primacy bias, where they choose the first answer that they agree with in a race to move through the survey. There also may be recency bias, where they choose the answer at the end of a list since it is more memorable.
  2. Question order bias – The order of the questions asked can create bias as well, especially as you seek more detailed information. If you move from more generalized questions to more specific questions, you tend to get a more accurate response. For example, if you begin with a broad question such as, “Were you satisfied with the product?” then ask about specific features and functions, you start from a general question and then ask the respondent to defend their initial response. If you reverse the order of the questions and start with more specific questions, such as “Did you like X feature?” then conclude with a question about overall satisfaction, you will get a different response based on the cumulative effect of the survey questions.

At last, we not only get to “why?” but we also get to “what can we do better?”

No matter what type of I survey you create, avoiding bias is always challenging. If you want uncensored feedback you have to use open-ended questions. However, you have to avoid bias in the way you form the question if you are seeking accurate customer experience feedback. For example, asking the question “What did you like best about your call center experience?” will yield a substantially different response than if you ask, “Was your call center experience satisfactory?” How you phrase the question will obviously skew the results.

Tell Me Why?

Too often companies become so focused on net promoter scores (NPS) and customer satisfaction ratings (CSAT) they forget to ask the most important question, Why?

Of course, you want to know if the customer is happy, would they buy again, would they recommend, and so on. Above all, you need to know why they are happy, disappointed, or gave the response they did. The only mechanism to get to “why” is an open-end survey question.

We have the technology to get to “why?” Real-time automated speech recognition and transcription allows us to collect and analyze open-end responses, including capturing sentiment-emotion data. Assessing qualitative customer responses in real-time is already revolutionizing call center response. Real-time analysis of open-ends provides the detailed data required for closed-end service recovery. At last, we not only get to “why?” but we also get to “what can we do better?”

There is no reason to throw away valuable data. Adding open-ends to post call center IVR surveys can provide invaluable and rich insight about customer experience, with three to five times the data you get from an SMS or an email survey. If you are really dedicated to improving call center performance and optimizing the customer experience, you need to give customers a chance to really tell you what they think.

Leave a Reply

Your email address will not be published. Required fields are marked *