Posted on: February 20, 2014
Tedde van Gelderen
Founder & President
What Usability Testing Does and Doesn’t Tell You
We at Akendi are increasingly asked to conduct usability tests with digital products in the later stages of product development. This means we come in to test at the stage where the wireframes are already worked out and the client is working on the coding and visual / brand design of the digital product.
So, we create our testing protocol, recruit end users and conduct the usability test. So far so good. Once we present the results, however, additional questions come up. Questions like: so can you show me what the redesigned page will look like now? Half of the users didn’t complete their task, was that because of the software testing environment? We wondered if users thought this was the actual product that they were looking for? Did the users get a good sense of the value of this digital product? Did they miss anything? Relevant questions, no doubt, but how does this relate to doing a usability test as a research technique?
What’s in a Name?
Ok, let’s step back for a minute. Usability testing, like any other UCD/HCI/UX technique, has a specific approach with specific – high value – outcomes. This technique has two terms in it: ‘usability’ and ‘testing’.
Let’s look at the first term: we’re testing the usability of a digital product (or other user experience). Usability has a well-established definition as well. It includes effectiveness (can I do with the product what I wanted to do, can I find the information), efficiency (how long does that take, how much effort do I have to put in to make it work) satisfaction (was the experience positive, engaging) and learnability (how long is reasonable for me to learn how to use this).
Let’s look at the second term: we’re testing. That means that the product is being examined, assessed, measured, and verified, to determine, objectively, whether or not it meets certain defined criteria. The outcome is a pass or fail, just like with a school test, a product safety test and a software quality assurance test. Testing can, and should, happen early. With a prototype of the product, on paper or in mock-ups or simple code / HTML. Testing is often done at specific times in the development process, for usability testing, it is preferably done when there is an early interaction design, but there are additional times to test later in the development process as well.
What Usability Testing is Not
So, when we’re doing usability testing that is what we’re testing: the usability, how easy is it to use this product. So what aren’t we testing? This usability testing thing doesn’t validate the product value. This is because of the low number of participants in a test project – typically anywhere between 8-18 people. This is also because of the context of the test setting: the setting doesn’t replicate all the different contexts a user would encounter in ordinary life. We’re asking each person to perform a task, not because they want to or based on a need they have identified, but because it’s part of the test. Even though you do get some anecdotal feedback that is valuable during the test sessions, these anecdotes are simply no basis for firm statements on the customer / user value of your product.
Usability testing is diagnostic, not prescriptive: it clearly tells you where the usability issues are but it doesn’t give you clear, accurate design solutions. Users are quite capable of telling you something doesn’t work according to their needs, expectations and logic, but after that it turns out to be a lot harder to come up with reasonable solutions for the less than ideal interaction they just had. You can’t expect a participant, who has just done a usability test, to switch quickly from analysis to good design solutions without knowing the project / product boundaries and design rationale that led to this user experience.
Unfortunately, what we see too often in presenting usability findings are knee-jerk reactions / quick fixes that seem the most obvious solution at first glance. In reality, product designers need time to assess usability results in the full context of the project / product boundaries. They are the people who are much better positioned to come up with reasonable fixes to experience flaws.
What usability testing does do is identify what is likely to go wrong with your user experience flows, functionality, page layouts and where the user needs are not met or understood. It is truly the software Quality Assurance of Experience Design, making sure we maximize the potential of a product or service experience and not let some interactions – i.e. mechanisms of reaching value – get in the way.
Tedde van Gelderen
Founder & President
Tedde infuses Akendi, its services and methodology with his strong belief that customer and user experience design must go beyond a singular product interface, service or content. It should become deeply rooted in an organization’s research and design processes, culture,and ultimately be reflected in their products and services. A graduate of Radboud University, the Netherlands in Cognitive Ergonomics, Tedde has more than two decades of experience in experience research, usability testing and experience design in both public and private sectors. Prior to founding Akendi, Tedde was a founding partner of Maskery & Associates in 2001. He has worked for companies including Nortel Networks, KPMG Management Consulting and Philips Design.
Sign up for our UX Blog
Don't miss an article! We'll notify you of each new post.