Just how long is too long for a survey? This is a critical question that we market researchers must ask ourselves when we design surveys. What makes this critical is that society is becoming (or has become) fixed on instant gratification.

The days of 30-60 minute surveys are long gone! Yet some panel companies insist on putting respondents through a process that amounts to torture. I have to believe that this is not at the insistence of the researcher, but at the request of the agency or end client. I often wonder if clients can find the time to sit down and faithfully answer the questionnaire they hope to put respondents through.

So, what is a good yardstick to follow? I believe 15 minutes is the longest a respondent should be expected to participate. If that is the longest one would expect to sit still, then I recommend keeping your surveys to 5 to 10 minutes in length. If you incorporate audio-visuals or interactive questions in your online survey design then you can push the time barrier, under the theory that higher levels of engagement will help to support completion rates.

Contrast this with an experience a colleague of mine recently had with his cell phone company. After a service call he received a text message stating that his opinion was valued and asking if he would participate in a series of four single question texts. All he had to do was reply to each text with a numerical response. The texts were timed over a 48 hour period so as not to seem intrusive.

What this means to researchers, regardless of whether we focus on consumer or B2B marketing research, is that we need to be diligent in our design practice and how we manage client expectations. I will argue it is better to go with a series of short surveys, each centered on a specific topic than to put participants through an endurance exercise where they are apt to lose focus, speed through or drop out altogether.

What say ye?