Katie’s article is a plea to the HCI community to ‘wake up’ and propose new usability metrics and adapt itself to 2016. I agree with some parts of the article. But, the question arises: are the current usability methods really obsolete? I’d say not yet, but they will be.
It’s plausible that some of our usability methods might not fit well in 2016. Where toddlers learn alphabets on iPads, teenagers have smartwatches as powerful as some computers, and adults depend on phones to find partners. The game has changed, and there is more to test than the methods could handle. We need methods to measure and analyze emotional responses, especially for the new ‘social’ products. For example, the popular dating app ‘Tinder’ where people look for partners to date with. When does the user experience end? Since the user's goal is to find a boyfriend or girlfriend, is it a poor user experience if the date doesn’t go well? When does the user experience end? Several such questions need answers.
Katie talks about new trends that widen the gap between usability theory and industry practice. One of them being “developers being designers” or “users being designers.” We shouldn’t complain about this because “designers need to think like business people.” While everybody is an expert in their domain, the skills borrowed from other domains ease quality work. Increasing accessibility of design tools with low learning curves has lead non-designers to explore design. They’re starting to take control of something they think they understand. There could be reasons for them to believe so: good access to early adopters that don’t care about novelty over experience or zero competition.
Designers advocate that usability should dictate whether a product is good or not. The users or even the businesses might disagree. For example, it’s more important for businesses to get the product out faster because that generally dictates who leads the market in that space. Many market leaders are often substandard products, but they were the first to get released(first-mover advantage).
Katie feels threatened because the absence of usability testing isn’t missed. It’s fair to be threatened. If we look at the number of beta products silicon valley releases, the iterations on features, and experience through successive releases and updates, it's clear that the tech industry has somehow figured out the alternative to usability testing. As much as we hate it, it’s working well for them at the moment.
She is worried that the loss of usability testing will lead to poor products. This is not true since poor products lead to poor products, not a process check like usability testing. Even if there is compulsory usability testing, the product managers will take the calls to address the said issues in that release.
The question directed towards us is: Is our future at stake as designers? If people are successful at making usable products without designers, is that good or bad? Designers will never be out of jobs. In fact, let people try. They will make terrible products and will realize that they need designers all the way more.
We even have to rename or redefine the term “usability.” It’s beyond “ease of using” the product. VSCO and Snapchat updated their applications to unconventional navigation patterns. Snapchat’s majority users base accepted the change. But, VSCO’s community dissed the change. Is VSCO’s update bad design but Snapchat’s is not? When do we take that calculated risk?
Complexity-wise, there are more variables now than our methods can handle. All these variables have different qualities. For example, requesting a cab on Uber. Some of the variables are the availability of cabs in the neighborhood, driver’s availability, their state of mind, mood, and current activity. Katie talks about the same in the article using Urbanspoon’s example.
Building on Uber’s example, let’s suppose the primary user is the person who wants a cab. The driver is the secondary user and will accept a cab request user. The experience for the primary user may depend on the behavior of the secondary user, and both users interact using devices. Their collective interactions in their respective contexts combined with devices' digital interactions with the servers over the internet all sum up to make a distinct character. Our methods sure do not incorporate all these factors in their analysis. But this is why beta tests are a good alternative because not all failures can be predicted. What if we have new usability methods to predict the issues that would surface out of beta releases? To answer this question, beta tests can be the ad hoc alternative to usability testing.
To sum it up, the applications that exist today are so different from what people imagined 10 years ago. And our usability methods are still the same. But, they are not outdated; they fall short of being applicable for all cases.
HCI should still care about usability. Because if we don’t, nobody else will. A few years from now, our products will evolve and become pervasive. Wearables might be the only personal device. How will we plot them on the usability scale? Instead of catching up on the time we’ve missed out on by not adapting the usability to strain current products, we can start exploring methods to perform usability on the technologies that are predicted to surface.