How much value can one extract from a questionnaire? I think of these documents as data generation cubes, and there are three types of data. One set is about the respondent – their demographics maybe, but also their attitudes and strategies towards the category in question.
The second set – Type B – is about the brands or competitors within the category in question. Rating their features, discovering our drivers, comparing their benefits etc.
Type C, or the third set of questions, typically, are the “end of the day’ questions. And also “how do we reach you?’ questions – such as media or social media consumption so that we can identify the likely customers, but also work out how to reach these golden people.
Now specific questionnaires will have their own topics, but for the most part those three main types of question just about cover it.
The secret to getting more value is to maintain something like a 33:33:33 ratio between the three types of question. If you had just 12 questions and asked 6 Type A questions, 4 Type B question and were left with just 2 Type C questions, then you’d basically get 6x4x2=48 units of data. Not bad. But if you asked 4x4x4 you’d get a lot more information: 64 units of data or 33% more information just by re-jigging your questionnaire.