A perfect user interface - the art of asking the right questions
Last night I watched a TED lecture by Malcolm Gladwell in which he honours Howard Moskowitz – a brilliant US market researcher, who has worked with an endless list of companies to understand their consumers’ needs. The talk highlights a study Moskowitz undertook for Pepsi to find the right level of sweetness in a new aspartame-based diet drink. Having conducted numerous focus groups, Moskowitz started plotting the collected data and realized that it did not look like a traditional bell-shaped curve but is, instead, sporadically dispersed. The revelation came later when he realized that they had not asked the right questions. Instead of looking for the right Pepsi, they should have looked for the right Pepsis!
This realization had sparked a revolution in food science and hence, food industry. The question they should have initially posed was about the variety of Pepsi drinks consumers would like to see on the market, not the sweetness level of one Pepsi. Restating this simple research question brought us, the consumers, the ultimate variety of Pepsi, and later, spaghetti sauce, as we see today – chunky, spicy, and plain. But the sauces aside, it brought us the variety!!!
The traditional approach of asking the consumers “Tell us what you’d like and we’ll give it to you” had been overturned. Apparently, we – people, cannot always explain what we like in precise terms. We tend to stick to collective universal concepts, to one-size-fits-all, while subconsciously we strive for variability. The Horowitz’ genius caught this evasive variability principle!
The art of great web design is founded on the principles of thorough knowledge of the target audience needs and building the interface around the anticipated user interaction with it. Not surprisingly, the same “Pepsis” precept discovered by Horowitz applies to UI research – ask the right questions!
Various techniques of interface evaluation, such as usability testing, surveys, interviews, focus groups and such, will you provide enough information to assess the efficiency of the UI. Asking very high level, open-ended questions, such as “What do you like about the interface?” will hardly provide tangible data for which you seek. Decomposing and slicing your questions to present objective options with a rating scale would provide measurable results. For example,
-the interface enables me to accomplish tasks more quickly (rate on a scale 1-5, strongly agree – strongly disagree)
-the navigation helps me to move around the interface effortlessly and quickly (rate on a scale 1-5, strongly agree – strongly disagree).
Even if these questions don’t provide answers to creating a new type of spaghetti sauce, at least you will have a great website!
