Open-ended Questions: Lazy or necessary?

I don’t hate open-ended survey questions, but I do hate lazy survey questions and oftentimes (but not always), open-ended = lazy.

You see, I’m all about people getting the best information possible from their surveys — that means they are getting quality information, and they are getting it quickly and easily.

Open-ended questions can be very challenging to analyze, and oftentimes they should quite obviously be closed-response. Let’s look at a quick example.

Below is an open-ended survey question from the clothing company, Reformation.

For starters, this question is not a question, it’s two. But alas, there is only one text box. So, people are maybe typing both answers into one box. Maybe they are only answering one of the two questions. Maybe they are just typing random letters. I don’t know.

I do know that Reformation is quite popular. And, if they are surveying everyone who makes a purchase on their website, they are looking at thousands if not tens of thousands of open-ended responses (also known as words or letters people have typed into the box provided).

Oh, how I feel for the poor analyst who has to go through those thousands upon thousands of text boxes and (1) parse out the answers to two different questions, and then (2) make sense of people’s answers to those questions. Sounds time-consuming. But who are we kidding, they probably aren’t going about this in a systematic way. (sad face).

Open-ended questions like this are bothersome, because they could so easily be a closed-response question (i.e., you provide options to choose from) that could be meaningfully analyzed within mere minutes.

Make it stand out

Whatever it is, the way you tell your story online can make all the difference.

Above, I just quickly pulled the items listed in their website’s navigation menu and made a simple yes/no follow-up. (I also spelled out their company’s full name — I don’t love it when survey’s try to be “cute.” Cute is confusing).

You could envision this being a question where people can select multiple items and then answer yes/no for each one. There is a lot you can do here. But notably, there is no good reason to have this be an open-ended question. Let’s not waste their analyst’s time or delay them getting important information like nobody is finding the shoes they want on their website :)

Good survey questions are designed to get you the information you want to know, and to get it in a quick and easy way.