From research to loyalty marketing: the many uses of telephone research
The telephone survives as a research mode because of what it can do that self-completion cannot: reach a named individual, probe an answer and control who actually responded.
Every few years the telephone is declared finished as a research mode, and every few years it turns out to be doing work nothing else does. The reason is not nostalgia. It is that a telephone interview is the cheapest way to be certain who you are talking to, and the only practical way to ask a follow-up question that was not written in advance.
What the mode is actually good at
Reaching a named individual. In business research the target is usually a role, not a household: the person who signs off the maintenance contract, the person who chose the supplier. Self-completion reaches an inbox and hopes. A telephone interviewer can navigate a switchboard, confirm the role and, if the wrong person answers, ask to be passed on.
Probing. A questionnaire can only anticipate. An interviewer can hear that an answer was hedged and ask why, which is how the unexpected finding — the one that changes the brief — usually arrives.
Controlling the response. In a self-completion survey you do not know who filled the form in, how long they took, or whether they read the question. On the telephone you do.
What CATI adds
Computer-assisted telephone interviewing puts the script on the interviewer’s screen and the answer straight into the dataset. The gains are mostly unglamorous and entirely real: routing and quotas are enforced by the software rather than by the interviewer remembering; the separate data-entry step disappears, and with it a class of transcription error; and the evidence suggests reduced interviewer variance, because everyone is reading the same thing in the same order.
Experimental comparisons between paper and CATI versions of the same instrument have found few differences in the response distributions themselves. The case for CATI is efficiency and control, not a change in what respondents say.
Beyond the survey
The same infrastructure does more than measure. A structured call can confirm that a service problem was resolved, reactivate a lapsed account, or qualify a lead — and the boundary matters enormously. Research and marketing calls have different rules and different duties to the person on the other end. A call that starts as research and ends as a sales approach is not market research, and the industry codes are explicit about it. Keep the two activities separate, and say which one is happening at the start of the call.
