How to Build an Effective Questionnaire for Competitive Intelligence Expert Interviews
A questionnaire to generate feedback is crucial to obtaining the information you need to develop competitive intelligence. Ask the right questions in the right way, and you’re likely to receive high-quality information from experts you send it to. Ask the wrong questions or use ambiguous wording, and the results are likely to fall short of the goal.
How to choose the right Q&A format for expert interviews
When developing a questionnaire, you have several Q&A formats from which to choose:
Likert scale questions: A Likert scale involves using an odd number of responses (5, 7, and 9 are most typical); for example:
In your opinion, how successful would you expect a chocoholic to be if promoted to management?
A. Highly successful
B. Very successful
C. Could go either way
D. Probably not successful
E. Definitely not successful
Use the Likert scale whenever possible. If properly designed, it covers the entire range of possible responses and is very useful when employing some form of the Delphi Method.
Forced response questions: A forced response question uses a scale that makes the interviewee indicate a preference or decision. Here’s an example:
What is your view of hiring people who are chocoholics (meaning they’re addicted to chocolate)?
A. I strongly recommend hiring them.
B. I moderately recommend hiring them.
C. I somewhat oppose hiring them.
D. I strongly oppose hiring them.
When composing forced-response questions, always include an even number of answer choices to discourage people from playing it safe and choosing the middle answer or the “no opinion” response. Many people opt for the least definitive, most noncommittal response. If you want more honest answers to a question, providing an even number of choices forces them to commit to one side or the other of an issue.
Obtaining honest feedback on emotional or sensitive issues is often difficult because people want to avoid expressing their true opinion on a topic. You should strongly consider using forced-response questions whenever the potential for respondent bias is high.
Open-ended questions: If you’re working with one expert or a small panel of experts, consider using open-ended questions. Open-ended questions are often the easiest to compose, so if you’re facing a time crunch, these questions can really come in handy.
Open-ended questions are most suitable when you want the respondent to express an expert opinion and lead you to new areas of knowledge. Just be prepared to ask follow-up questions. Think of your questionnaire as a conversation starter. After you’ve had time to review each expert’s answers, the next step is to discuss their answers with them in person (or via videoconferencing) so you can really go in-depth with them.
Closed questions: As opposed to open-ended questions, closed questions provide a limited number of choices — usually yes or no or true or false. You may include a third option, such as “Don’t know,” to let the interviewee opt out.
Closed questions are useful if you want to know someone’s opinion within the constraints of two or three choices.
Multiple-choice questions: This type of question enables you to give an interviewee several answer options within a given range. Most pollsters use multiple-choice questions that allow for answers from one extreme to the other. However, they can also be used to compare, say, different brands. Here’s an example:
Which of the following best describes your view of the future of WiMax?
A. It will be the technology of the future.
B. It will continue to have moderate success.
C. Its future looks marginal.
D. It will continue to decline.
E. It’s a long-term loser; no future.
Use multiple-choice questions to poll larger groups of experts.
How to compose questions for expert interviews
Writing questions that elicit the most accurate information and insight is more art than science, but here are some guidelines to get you started:
Be clear and concise. Write your questions as terse statements and then convert the statements to questions. Wordy questions are likely to either bore or confuse respondents, neither of which is good.
Use concrete language and answer choices. For example, asking someone if she would recommend a certain technology is more likely to elicit an honest response than if you ask whether the person likes the technology.
Avoid using vague words or introducing ambiguous concepts.
Cover all choices. When presenting a list of choices, make sure you cover all options; otherwise, you’re likely to either receive some unanswered questions or have people choose what they think is the closest match.
Avoid any biased words or phrases. You don’t want to disclose any information that may reveal your opinion about an issue in question. For example, instead of asking if the respondent supports increased government regulation in the industry (a question that’s likely to elicit a negative response), ask what the person thinks would be the likely result if a certain regulation were implemented.
Always try your best to eliminate bias from questions and the way you ask them.
Mix broader questions with those that are more specific. This helps to obfuscate (to the respondent) the true nature of your line of inquiry. Here’s an example of two questions that would work well together because they prevent the expert from guessing why you’re asking these questions:
What are the pros and cons of XYZ technology?
Which company is leading the pack in XYZ technology?
Try to hide the most significant questions in the midst of a more mundane series of questions so the person completing the questionnaire is more relaxed and open when encountering the big questions.
Arrange questions logically. You can arrange questions chronologically, from least to most complex (or vice versa), from less to more sensitive, from general to specific, and so on. Consider grouping related questions to prevent respondents from having to shift from one topic to another.