Generating IVR Invitations

Interactive Voice Response (IVR) surveys are becoming a popular methodology for less expensive large-scale customer satisfaction surveys. To make the results from these surveys statistically accurate, customers must be selected at random to participate, and the most significant obstacle often becomes the ability to randomly generate survey invitations. One common automated approach is randomly generating survey invitations on register receipts on an Nth basis at the point-of-sale. What does a company do when they don't have that capability?�

The primary concern when a company does not have an automated means of selecting potential respondents is that as employees need to become involved in the invitation process, they might cheat and only invite those customers they think will give highly satisfied ratings. This is known as gaming. When computer-automation is not available to take the invitation process out of the employee's hands, then the next best thing you can do is establish strict parameters for participation that all employees are required to follow.�

We'll look at two examples - a call center with a PBX system that does not have the capability of randomly selecting callers to transfer into the IVR survey system, and a chain of stores that does not have the capability of randomly generating IVR survey invitations at the point-of-sale.

With the call center, invitations can be extended on selected days and hours for all service representatives (or a portion). At the end of customer calls during that time period, all callers are asked if they want to be transferred to the survey. Your marketing research partner can provide the number of actual call transfers during that period and help you compare transferred calls (not all will finish the survey) to completed service calls -- establishing a traceable transfer percentage. The combination of monitoring service calls during the survey period to assure all customers are being invited to participate and monitoring transfer rates has proven very effective. With sufficient monitoring, cheating can be held to a minimum.

With chain stores, all customers on selected days and hours are given an IVR survey invitation postcard containing the store code at the point-of-sale, and are verbally asked to participate in the survey. Again, store managers and chain executives should conduct spot checks of the stores to ensure that all customers receive invitations during the survey period, including those who do not appear to be happy with their shopping experiences. As with call centers, stores will build typical response rates over time that can be tracked.

All customers must have an equal chance of being selected if the results of the survey are going to be projected onto the entire customer base. To further ensure that a wide variety of customers are included in the survey, survey days and times should vary in association with call center or store hours -- taking customer volumes into consideration. Since customer volume, response rates and transfer rates vary, it is always a good idea to start IVR programs with a smaller pilot-test to confirm your assumptions. But with careful planning, it is very possible to overcome obstacles associated with the inability to randomly generate IVR survey invitations.

All Content � Copyright 2004-2011. Polaris Marketing Research, Inc. All Rights Reserved.

Send inquiries to or call 1-888-816- 8700.