Don't call us, we'll call you

Conventional wisdom says that most consumers don't want to be bothered to participate in a research study. Well, that notion is being challenged by an increasingly popular technique in which the respondents are the ones who call the researcher.

In this case, the "researcher" happens to be a machine, a sophisticated software program that lets respondents use their touch-tone telephones to take an "in-bound" survey. Their impersonal nature, and the fact that the respondent controls when the interview is conducted, may be part of the reason in-bound surveys earn better than average response rates for some users.

One firm that has successfully used in-bound surveys is Long John Silver's Restaurants, a quick-service seafood chain based in Lexington, Ky., with nearly 1,500 locations in 38 states, Canada, Singapore and Saudi Arabia. Long John Silver's uses Show N Tel, an in-bound survey program made by Interactive Communications Inc., Dallas, that runs on an interactive voice response system.

As part of its regular research, the company last year used the in-bound surveys to reach an elusive customer segment - drive-though and carry-out patrons. "A big part of our business is drive-through and carry-out customers," says Larry Noble, marketing research manager, Long John Silver's Restaurants. "While it's easy enough to talk to our dine-in customers it can be difficult to get drive-through and carry-out customers. [The in-bound survey] has worked the best for us in reaching these people. Up until we did the surveys, we didn't have a lot of information on those customers. We've tried mail-in surveys but the return rate is historically low."

Cards were distributed to Long John Silver's carry-out and drive-through customers inviting them to call a toll-free number by midnight to participate in a short survey. In exchange they would be given a code at the end of the interview which validated the card as a coupon they could redeem for a free meal and beverage of their choice.

When the respondents called in, they were greeted by the system: "Hello and thank you for calling the Long John Silver's Survey Line. We're interested in your opinions regarding your recent visit to our restaurant." Once it was established that the caller was using a touch-tone phone, he or she was told how long the survey would take and how to receive their incentive.

While Noble won't elaborate on how the results were used, he says that the information gleaned from the surveys was valuable. "The main advantage [of the in-bound approach] was that it helped us get in touch with a customer segment that we hadn't been able to reach before as effectively. When the need comes up to talk to those people again we will use it because it worked well for us."

Immediate responses

Mark Mulch, national project director, Interactive Communications Inc., says that one of the benefits of the in-bound approach is the immediacy of the responses. For example, instead of recalling a visit they made to a restaurant days or weeks ago, they're quite often responding to something they did hours ago. "We're getting responses from people who have just been to that restaurant that day. They respond very, very quickly while the experiences are still fresh in their minds," Mulch says.

Long John Silver's customers were asked about the food items they purchased and given a rating scale to register their satisfaction on a number of attributes. Respondents also provided standard demographic information.

Mulch says that three types of question work well in an interactive survey: rating questions, in which respondents press a key corresponding to the correct number on the scale; data entry questions, such as "How many minutes did it take for your order to be filled?"; and open-ended questions, where respondents explain why they weren't satisfied.

An in-bound survey can contain any number of skip patterns, so that if a respondent rates a product poorly, for example, the program will take him or her to an open-ended question to probe their thoughts further, with a request such as, "Please tell us in your own words why you rated the product so low."
Respondents to the open-ended questions aren't told how much time they have to respond but Mulch reports that 45 seconds is typically enough time. Once they are through they press a key on the phone.
Clients receive the open-ended responses in transcribed form or on a cassette tape. The system also allows clients to call in and listen to the verbatim responses. So, for example, a store manager or an area manager can call up and get a quick read on customer satisfaction.

Safeguards

The system includes safeguards to prevent duplicate responses, Mulch says. "We can put a control device into the survey - for example, having respondents enter their phone number. If someone with that number has already called in, the computer would say 'I'm sorry we've already received a response from your household. Thanks for calling,' and give them validation code anyway."

The system also can keep track of responses by store, allowing for limits on the number of responses from patrons of one location. "When the respondent calls in, they enter the store number and that raises the counter on that store so once they hit the target number the database cuts it off and says 'We're sorry, we've completed the interviews for this store.' Even though they've reached their quota, the respondent still gets the validation number to get their free meal, to reward them for going to the trouble of calling."

Strength and weakness

The absence of a human interviewer is both a strength and a weakness of in-bound surveys. Because respondents are interacting with a computer, their responses may be more honest, particularly to the open-ended questions, because they aren't worried about offending the interviewer. In addition, they may feel less hesitant about revealing sensitive data such as income levels.

Another benefit of the computer-as-interviewer is that every respondent hears every question in the same way. "In a telephone survey, how the interviewer asks a question at the beginning of their shift may be different than how they ask it at the end. But the computer never has a bad day," Mulch says.

On the flip side, in-bound surveys lack the personal touch of live interviewers, most importantly their ability to probe for more information. But that drawback can be overcome with a little qualitative research. If the in-bound surveys uncover problems, they can be more thoroughly examined through focus groups or one-on-ones.

Positive impression

While in-bound surveys aren't likely to supplant good old-fashioned telephone interviews, as the Long John Silver's example shows, they offer an efficient way to talk to hard-to-reach customers. Their anonymity may also make it easier to gather sensitive information such as employee opinions. They also may show consumers that participating in research can be interesting and fun. After all, if a respondent goes away from an in-bound survey experience with a positive impression of the research process, they may be more receptive next time a researcher calls.