Editor's note: Chris Van Derveer is president of Van Derveer Industrial Research, New York City.

Many marketing managers face a common dilemma at the conclusion of a strategy planning session for their companies. They develop, or are presented with a requirement to conduct, market research. And while the company may understand the specifics of the research data it requires, it is left up to the marketing manager to select the best methodology for gathering that data.

Not only must the most accurate method of managing the work be selected but the chosen method must conform to a strict budget for the work as outlined by management. The purpose of this article is to guide industrial research managers through the four major methodologies: telephone and mail surveys, focus groups, and personal interviews. Though these methods are also used in consumer research, there are vast differences in how questions are asked and, of course, industrial questionnaires never ask lifestyle questions, as they have no direct effect on product purchasing.

Telephone surveys

Telephone interviewing is by far the most popular methodology used in industrial marketing research. The method is quite rapid - large populations may be surveyed and critical market data may be returned to your offices within weeks after surveying has begun - and it's accurate, because pre- screening procedures ensure that the person most knowledgeable about the survey topics is interviewed. But telephone surveys can be limiting due to their length - usually 12 to 15 minutes of a manager's time - and the fact that great detail cannot be developed from them.

Also, it's a waste of time to describe the visuals or layout of products through phone research. If designs are discussed, they must be pre-mailed to respondents and be in front of them at the time of the survey; a more expensive process indeed! Telephone surveys are always much less expensive than personal interviews and are slightly more costly than mail surveys, of course depending on how those mail surveys are managed.

It's best to use telephone when you require a rapid turnaround, do not need highly detailed data, and need data which is projectable. In justifying management decisions, which may involve investing millions of dollars, the data that you base your plans on must reflect the real statistics of the audience. If you conduct SO interviews, for example, to gauge the opinions of 6,000 electrical distributors, you could develop data that is highly inaccurate. As a matter of fact the final result may vary +/-15-20 percent from any mean that is developed in the project. This variance is not acceptable for planning purposes. Any decision based upon it could result in a marketing disaster.

Real, projectable samples begin at a sample size of 100 interviews (a +/- variance of 10 percent). For most industrial projects a sample size of 300 interviews within a specific market segment will return a result with a variance from the mean of +/- 5 percent. In industrial research several hundred surveys will usually achieve your goals as the number of industrial purchasers may number in thousands, as opposed to consumer goods buyers, who may number in the tens of millions.

Mail surveys

This technique comes and goes in popularity. I believe it is currently out of favor, but as the winds change so may its standing. Companies with limited budgets for research often turn to mail surveys because the initial outlay is small and they may have used them in the past. These firms will always get data back but I have always questioned the accuracy of what is resumed to them.

Mail surveys are erroneously believed to be much less expensive than telephone. A mail survey must gain a 50 percent level of return to develop data which is accurate to measure the statistics of the audience. The reason a 50 percent return is required is due to the inherent inaccuracies of the process. The major problem is the quality of the response. When the survey arrives in the office, anyone can fill it out and return it. By contrast, telephone respondents are pre-screened.

In terms of expense, the initial mailing may, at first, be cheaper than a telephone survey. However, if your first wave of responses is only 20 percent of the group, you must then continue to send several waves to get up to 50 percent of the sample. By the time you have conducted your third wave and taken about six months to do it, you will have spent as much or more on a per survey complete basis for mail vs. telephone.

It's best to use mail when your group has the time to wait and you want to develop data that is more detailed than the data you can get from a telephone survey. Depending on the patience of your audience, mail questionnaires can be many pages in length. In addition, product designs can be pictured and rated - something you can't do with telephone.

Typically, I would recommend against mail for all of the inaccuracies that it has. If you have used mail research in the past, change to telephone by trimming the number of questions and by scripting them in a tele-conversant format. If you are receiving a 50 percent return level, and you know through experience that you are getting the best respondents and must cover many topics, by all means continue to use mail.

Focus groups

The most important thing to keep in mind about focus groups is that they are not samples that are projectable to an overall market audience. This is only a qualitative technique and is used as a way of conceptualizing issues, not exactly measuring them.

A group is about two hours in length and usually consists of up to 10 qualified and pre-screened respondents. Focus groups are simply structured discussion sessions where companies question informed users about specific topics and/or product requirements. Clients often ask me how many focus groups are necessary to evaluate qualitative issues for one market segment. I usually recommend three, because if we conduct three groups we can better search for a consensus of opinion. Focus groups are especially useful in defining issues so that they are accurately presented on telephone or mail survey projects that may follow the groups.

It's best to use focus groups when you do not fully understand the issues facing you in a new market, where the topics evaluated may be complex, and where group interaction may bring out opinions better than other techniques. This concept of group interaction is the key to success in this methodology. Not only can detailed opinions be gained, but when respondents are allowed to interact, new and creative insights may be developed. Typically, telephone or mail surveys, which are very scripted, do not challenge the respondent to creative heights. As I mentioned, take what you learn in the focus groups to design the quantitative research that follows.

Personal interviews

Compared to the other three techniques, personal interviews are by far the most expensive to conduct on a per survey complete basis. The major part of this expense, of course, is the cost of traveling to far flung interview sites. However, for specialized applications, personal interviews may be the only option for developing high quality data.

It's best to use personal interviews when the nature of the data you require is detailed and you have a large budget. Typically, the interviews last up to an hour and can be used to cover scientific, financial, engineering or other complex topics that cannot be well evaluated through the other three methods. The technique is especially well suited to visuals or display of products for rating of their features and abilities. For the sake of credibility, it's important that the interviewer you use is a technical expert on the topic.

Finally, personal interviews, regardless of the survey topic, are many times conducted with high profile executives who can't be reached through other survey means. This interview technique is more respected by executives and as such they may accommodate it while rejecting other methods. Keep in mind that many executives require hefty incentives to be encouraged into the interview process.

This article, of course, can't cover every aspect of why each method may be best to meet your requirements for data. But by employing the general tips we've discussed, you can get your industrial research off to a solid start. For more information call me at 800-531-9025.