Editor's note: E.B. Feltser is a freelance writer based in San Diego. She has worked as a marketing research interviewer and survey writer.

Okay, so you're a telephone interviewer at the local friendly marketing research company. You're staring at the first screen of a new survey and waiting for someone to answer the phone. In the briefing, you were strongly reminded of the Golden Rule: Read The Survey Verbatim! And there on the screen is the intro. Is it strong enough for a verbatim reading? Or will you have to make changes in order to get the job done?

[ASK TO SPEAK TO THE MALE/FEMALE HEAD OF THE HOUSEHOLD. WHEN HE/SHE IS ON THE LINE CONTINUE.]

Hello. My name is Perky Interviewer and I'm calling from Hot Shot Surveys, a marketing research company. We're conducting an opinion survey about some topics of interest. We only want your opinions. There is no selling involved. Do you have a few minutes to answer some questions?

Sounds like a reasonable start, doesn't it? And it, or a close cousin, is certainly common. But in fact it's weak, and since continued employment as an interviewer is directly related to the number of completes, there's a strong, often irresistible, impulse to make changes. After all, a telephone interviewer has maybe 10 seconds to hook the respondent, and the only tools available are the words of the survey and the interviewer's vocal persuasion.

So, what's wrong with the words? Lots. Look at it point by point.

  • Speak to the male/female head of the household

Anybody who's ever juggled their bill paying knows that phone calls that start out asking for the head of the household are bad-news calls. Predictably, this opening seems to bring on a rash of respondents who claim to be baby-sitting for people who work erratic schedules 12 hours a day, seven days a week but, "Sure, call back tomorrow and maybe you'll catch them in." Apart from shooting you down before you're barely off the ground, this sort of initial-refusal-disguised-as-RNA also clutters up the call-back pool. Move this request to the end of the intro.

  •  A marketing research company

Okay, you and I know the difference, and we're proud to be working in the wonderful world of marketing research. But "marketing research" sounds like "telemarketing" to a whole bunch of people out there, and if you're lucky they won't snarl at you before they hang up. One easy fix is to change "marketing research company" to "opinion research company" and thereby nip the problem in the bud. If this is too much like a betrayal of your ideals, use "marketing research" but follow it fast with the disclaimer, "I'm not selling anything. We just do opinion surveys."

  •  About some topics of interest

This sort of vague phrase prompts many a respondent to ask, "What topics?" The interviewer must choose how to respond. Either make up an answer and risk blowing the survey design, or repeat the vague phrase and risk coming across as evasive and even downright sneaky. It's much more helpful when this what-it's-about statement is as specific as possible within study-design parameters: about household cleaning products, about automobiles. Adding geographic scope is always helpful, because it adds the weight of serious intent: a national survey about. . . ; local issues of education.

  • We only want your opinions. There is no selling involved.

This disclaimer should be brief, personal and in the active voice. I'm not selling anything, we just do opinion surveys. Consider what it actually says: "I understand your disgust about telephone sales pitches and I wouldn't dream of doing that to you. I just want you to tell me what you think about some stuff." It's the first opportunity to connect with the respondent at a human level and start constructing the bond of trust and tolerance that leads to a completed survey. This holds true whether an interviewer favors the "I am a disinterested computer" camp's no-nonsense approach or the "I am a warm, trustworthy human being" camp's chatty approach.

  •  Do you have a few minutes?

In the ongoing debate about whether to ask this, my position is a firm "it depends." If it's mentioned, respondents frequently want to know how many (read: how few) minutes, and the survey language must supply an answer. On short surveys interviewers will often just get on with it and not mention times, although a short time frame can be an asset ("It's really short, only about 10 minutes.") But if you expect respondents to spend an unpaid half hour answering questions, at least warn them so they can turn down the heat under dinner, or whatever. One graceful way to break the news is with words like "detailed" or "comprehensive," as in, "Well, it's pretty comprehensive. But most people finish in about XX minutes," with five minutes knocked off for every 15 minutes of normal running time. The "most people finish in" phrase is also helpful, since respondents usually figure they'll do it faster and when they don't, the burden's on them.

The bottom line is, if you want your intro to be read verbatim, make sure it's something interviewers will actually read verbatim. Read it aloud, and listen to it. Does it use language that real people use? Are there reasonable answers to predictable respondent questions, such as topic and length? Finally, is the intro interesting enough to tempt you to donate part of your evening to answering the questions?

Take the time to write a strong, winning intro. Those first few seconds are crucial, because if you can't get past the intro, you can't get the complete. Besides, interviewers will instinctively start to "strengthen" a weak intro, and who knows what that might do to your survey design?