9-27-12 Making sense from an informal, open-end survey

Readers of this Blog know I have strong feelings about doing legitimate statistically valid surveys.  Still, many businesses choose to create simple questionnaires they distribute to customers to get a reading on products and/or services being offered.

A simple questionnaire is one that includes three or four open-end questions.  “Open-end” means that no possible responses are given to the person, so they can fill in anything they want for a response.

Trouble is, when you get a lot of responses, it can be difficult to assess a trend or consensus that you can use to improve or change what you’re doing.  So, here is the process I use:

Write down verbatim each of the responses for a question, going through about 20 completed questionnaires.  Then look at them and see if there are patterns that you can paraphrase, such as types of products, emotional responses (they like or don’t like something), general ideas or suggestions that fit a pattern, etc.  Once you can identify three to five items that seem to apply to many of the responses, you can assign a code number to each of these items.  As you go through additional questionnaires, you will hopefully see that most responses fit into one of the codes you have created; if they don’t you can create an “Other” category.  However, if you suddenly notice a pattern in some of the “Other” responses, you can combine them and create an additional code.

When you are done coding the responses, you can tabulate them and see what patterns emerge.

By the way, most people love answering open-end questions because they feel they can truly communicate something that is important.  However, open-end questions can be very frustrating to tabulate and come to some conclusion that will benefit a marketing program….unless you can do what is described above and create a process where responses can be tabulated.