Consider these poor practices for a first date conversation (or for a survey)...
One of the most frequent questions I have been asked over the past 20 years is “what is the BEST number of questions to put on a survey?” My answer to this question is the same answer I give for many other survey construction concerns: treat your survey like it is a first date.
If you were going out on a date with someone for the first time, how would you want the conversation to go? The successful conversation would progress from topic to topic, asking relevant questions. You would show true interest by asking your date to clarify and expand on what was said, not monopolizing the conversation or asking personal details too quickly.
Contrast that with conversations you have had with your worst date ever - which are similar to how many surveys are worded. Consider these poor practices for a first date conversation (or for a survey):
1. Monopolize the conversation and pepper your date with questions because hey, you never know, this might be the last time you see your date.
That is how many surveys are written: as if it is the last opportunity you will ever get to talk to your customers. I always advise organizations to focus their survey on one or two topics and save other topics for the NEXT survey. But few want their questions asked later; each department wants their questions asked now and the survey then becomes unbearably long.
Instead, the survey should be a conversation where you ask a couple of questions and the respondent answers and then you close the loop by later explaining to the respondent how their feedback was used to make improvements. Follow this rule and you will have the opportunity for a second date and can ask more questions.
2. Wait until the end of the night to ask your date a series of personal questions.
How often have we seen questions on gender, race, age, job title and salary income, no matter how irrelevant to the purpose of the survey. Worse, these questions are usually found at the end of the survey (“so as not to skew the results” they say). At that point the respondent often wonders if her responses are now going to be filtered or interpreted in light of knowing she is a female Hispanic in a decision-making capacity earning $75,000-$99,999. If not able to opt-out of the demographic questions, many respondents simply end the date right there by closing their browser.
Is this what you would do on a first date: wait until the date has ended then blurt out “oh by the way, how old are you and how much money do you make and where do you work?” Not unless you wanted a second date! Don't be creepy on your date or your surveys.
3. Ask your date a long series of multiple choice questions in rapid succession, not allowing your date to clarify answers.
That is what surveys often do. You may have wanted to rate the service somewhere between 4 and 5 stars but there is no option for this nor any comment boxes so you can better explain your thoughts. In the days of paper surveys, respondents went around this obstacle by filling in the margins or the back of the sheet to clarify their rationale for certain responses. But this option disappears in the online world if targeted comment boxes are not used.
Do not just ask your date Yes/No questions like it is an interrogation. Give your date the opportunity to talk and tell you more!
4. Make your date stammer and look silly by asking irrelevant or confusing questions
Do not exasperate your date. If you should already know the answers to “what is your gender” or “degree earned,” do not ask this on your survey – your date will know that you are not listening. Do not use complex branching in your surveys when a more straightforward approach will do. Do not ask survey questions that have nothing to do with the conversation topic or you will drive your date away.
Treating your customer like a first date with your survey helps fix so many errors. Doing so would nearly always eliminate these problems discussed in other lessons: (a) surveys created by committee where everyone wants to get their questions asked on the same survey; (b) mixing purposes; (c) having no focus or lack of purpose; (d) using inflexible vendor forms; (e) too much complexity; and (f) asking items for which you should know the answers.
Takeaway: Do not scare off your date! If you were verbalizing these questions to the respondent in person, do you think your “date” would lose interest before the end? Focus on clear topics, of appropriate length, and allow the respondent to express opinions with plenty of comment boxes.