Posted on: March 5, 2015
Why Iteration is Critical to User Centred Design
At Akendi, we firmly believe in the importance of engaging the user throughout the design process.The earlier, the better—better insight, better requirements, better design and less time, effort and money spent fixing things that could go wrong later.
This is a key characteristic of User-Centred Design methodology and what our trademark Experience Thinking framework is based on. In these practices, feedback and iteration play a critical role in ultimately creating a better experience for the user.
Before exploring this further, a brief review of the User-Centred Design process might be helpful. User-Centred Design can be sliced up many ways, but along with engaging users early and often, it generally has three phases: Analysis, Design and Evaluation.
The Analysis phase of the User-Centred Design process is really what we like to call early user research. User research can include, but is not limited to, ethnographic research, shadowing and user interviews. These methods help us more accurately identify user needs and the context of use. What is key here is not to confuse user research with customer research (for more on this, see A tale of two tactics: User research and customer research).
Investing time early on in the design process to really understand who your users are, what they do, and what they need is what helps create the framework for your design, i.e. the requirements. Unfortunately, even though “user experience” is becoming a more ubiquitous term, the importance of early user research that generates requirements for design is still poorly understood and rarely emphasized (see UX Insight: Measure Twice, Cut Once).
At Akendi, we like to think about design as having three distinct phases of its own: Information Design, Interaction Design and Visual Design. Information design is the first step in the process and involves putting all those requirements into an Information Architecture (IA) and functionality that meets the user needs. This is followed by Interaction Design, which involves creating paper sketches, wireframes and interactive prototypes—the focus here is on getting the navigation right. The final stage is Visual Design, where the colours, fonts, icons and all that other fancy stuff comes into play.
Evaluation is exactly what it sounds like: testing your design and all of the assumptions you’ve made up until this point. At Akendi, we follow a well-defined process for carrying out Usability Testing that considers four factors: Effectiveness, Efficiency, Satisfaction, and Learnability. We test with real users in a controlled setting and recommend carrying out Usability Testing with all fidelities of the design.
Testing at earlier levels of fidelity is a best practice that we highly endorse (in addition to early user research) and one that will help save time, effort and money. Why? Because if you figure out that your IA, for example, needs to be more flat because users have trouble finding things, you can change it before you have even started the process of wire framing, let alone visual design. I’d like to point out here that Usability Testing is an evaluation method, in that it will tell you what works and what does not, but it may not tell you how to fix it (see What Usability Testing Does and Doesn’t Tell You).
Repeat & Iterate
Okay, so let’s say you have followed all the right processes. Users have been engaged throughout and requirements have been formulated with early user research. You’ve done card sorting, again with users, in order to determine the best IA. From there, you’ve created your wireframes and have completed a round of usability testing. Your design should be pretty close to perfect, right? Well…not quite. Even the expert, tenured designer, following a flawless process and the best usability heuristics cannot guarantee a perfect interface and this is why I’m “re-iterating” that iteration is a vital component of User Centred Design.
It’s no surprise that iterating early can have a major impact on usability at minimal cost. If you test the findability of your IA using tree testing, you can iterate on the IA before even getting to the design. You can even iterate on your user research. Let’s say you conduct a handful of user interviews and you are starting to see some surprising trends. You can improve on that research by creating a larger scale survey to confirm your findings. Major usability issues are usually found in the first few iterations and improvements based on these iterations have the highest usability gains. So why not iterate early, fix things early and save time, money and effort?
What happens if you don’t iterate early? Let’s say, for some reason, it just wasn’t doable. The visual design has already been put in place and through one round of usability testing, it has been determined that the IA could be improved. What to do next can spark a lot of debate. While I would advocate for an iteration, I know some would disagree. Often times, egos, budgets and timelines can get in the way. It is definitely true that it will cost more money to make changes at this stage; it may also be difficult to convince anyone else to make changes because perhaps they’ve hired you as the usability expert to get it right the first time. As the perceived expert, it can be hard to explain that “we didn’t get it right the first time.” At this stage, however, the design still has not gone into development – there is still a benefit to iterating now and making the changes before the product goes into market. You may not win that battle however, and there may be no right answer here.
Going forward though, what we can do is emphasize the important role that iteration plays in User Centred Design and of course, be advocates for early iteration.