Want better UX design? Just ask.

We recently began work with a client that was looking for an ordering system for their customers. They have a tight group of current customers who still place orders without the help of an online ordering system and have been doing so for ages. While working on discovery and scope of this project, we found ourselves making a lot of guesses about whether or not an online ordering system would be something their clients would even want to use.

Rather than basing our scope and use cases on hunches and pure speculation, we resorted to that tried-and-true information gathering method: the questionnaire. We wrote up a brief open-ended survey to give to a list of customers supplied by the client. It’s one of those things that you always kind of want to do at some point on every project, but getting the chance to do it this early in the process proved to be invaluable.

During the first call alone, we gleaned information that we never would have thought of by just looking out from our sterile studio bubble. One customer after another reaffirmed some of our thoughts, thoroughly dismissed others, and consistently gave us new insights as we went forward with the project.

In the end, we were able to provide our findings to the client in an organized presentation. Along with giving their team a number of insights into their customers’ perceptions of the company, they got hard proof that the customers absolutely wanted to be ordering online. With new information in hand, we revised the project’s scope, rethought how we’d approach the implementation of the ordering system, and reworked designs based on some of the unprompted feedback we received.


Some things to consider when creating a survey:

Have an idea of how you want the project to go

There already needs to be some kind of thought as to how the project is going to look or work. Use what you know to rough out an idea and base your questions on that.

Think through questions and keep them open when possible.

Like in the beginning design process, you need to think of all the possible scenarios under which a project will be used. Base your questions on these assumptions, but steer toward open-ended questions rather than multiple choice.

Give the person time to give free thoughts.

Along the lines of open ended questions, there should be questions that will let them simply tell you things that they have observed in their everyday dealings with the company, both positive and negative.

Take time to revise your thoughts and even the questions while the research is on-going.

This isn’t a peer-reviewed project. While the questions should remain FAIRLY consistent for the sake of comparison, if you’re finding a few questions just don’t work, feel free to drop them. Likewise, if people are bringing up topics you think are important, ask others about those same things.

Keep it short.

People are busy. 10-15 minutes tops. Also, phone calls are a nice middle ground between a written email survey and an in-person interview. Written questions often take too long to answer and it doesn’t give you the opportunity to change questions on the fly based on feedback. In-person interviews simply take too long considering setup and travel time. Even for a very busy person, it’s easy to find 15 minutes to chat on the phone.

Let them talk about them.

Again, feel free to go “off-script” especially if it lets the subject steer the conversation. People love to have an engaged listener, especially one that has an ultimate goal of making their life easier.

Take notes when doing the questioning.

We used spreadsheet that had a cell for each question and two additional cells for “Big ideas” and “Additional takeaways.” We jotted down notes about each response and created a short summary under our “other” columns. You don’t need exact quotes, just the gist of the each response.

While having an organized list of questions and responses will help get a complete picture, it’s important to know that since this is an informal survey (likely without enough responses to be technically considered statistically significant) you don’t have to be perfect in your questioning and collection of answers. More than anything, you should focus on getting a more broad, general idea of what the subject is saying rather than worrying too much about the formality of the survey. Don’t worry if you didn’t get an answer for one or two of the questions. If you can walk away with a general idea of the feelings of each individual subject as well as the broad opinions of the group as a whole, you’re going to have a good grip on the data you need to successfully inform a project.