In Designer Duds, Mills Baker talks about design that ignores people’s needs.

And if our best designers, ensconced in their labs with world-class teams, cannot reliably produce successful products, we should admit to ourselves that perhaps so-called “design science” remains much less developed than computer science, and that we’d do well to stay humble despite our rising stature. Design’s new prominence means that design’s failures have ever-greater visibility. Having the integrity and introspective accuracy to distinguish what one likes from what is good, useful, meaningful is vital; we do not work for ourselves but for our users. What do they want? What do they need? From what will they benefit? While answering these questions, we should hew to data, be intuitive about our users and their needs, and subject our designs to significant criticism and use before validating them.

It is a well written essay that raises some great points, but I disagree with the concluding thoughts. The conclusion rests on the idea that the user knows what they want, and that what they think they want is what should be pursued. This may sometimes be true, but it is also the case that oftentimes the users don’t know what they want. In cases where the users think they know what they want, their ideas about achieving it may not be optimal. Additionally, the needs and desires of the users aren’t born in a vacuum but are themselves the product of their environment – that is, they are the product of cultural and societal norms, or, simply put, the product of other people asserting what they think is important.

As such, putting apparent user needs at the top of the design process transforms it into something like a mirror that simply reflects what is around it but does not itself lead. That would be well and good if the objects of its reflection were unquestionably optimal, but, for the reasons above, that is not the case.

Relatively small interface improvements can be chained together to create leaps in user experience. The original iPhone chained together three such improvements: 1) finger touch interface to replace the stylus, 2) capacitive sensors instead of resistive, and 3) elastic scrolling effects. Touch screens have existed for a long time before the iPhone, but their interfaces, often stylus based, generated a lot of friction. That friction was removed by implementing the three interface improvements above, resulting in something that felt smooth, natural and pleasant to use.

You could not ask your users whether or not they wanted this thing because this isn’t something they knew at the time. Nobody has used such a touch screen before, so few people had an idea of the sort of experience it would deliver. Designers had to trust their own instincts, had to trust their vision. Your users will only tell you whether they like something or not after they use it, but this is not possible when the thing doesn’t yet exist. The role of the designer is to tell whether something is a good idea or not before it takes shape, before it is fully realized. They can do this not because they possess an innate gift, but because they have enough experience to recognize successful patterns, enough experience to imagine how these patterns would interact together in the final product. A good designer can see and experience their product in their mind just as a musician can hear their composition in their head.

An innovative design does not make a project a success or a failure, and I am certainly not arguing for some intrinsic value of design that elevates a clever interface above other metrics of value. A design that fails to deliver to its intended user base has failed. There is no way of getting around it, and it is not my intention to do so. My point is that an experimental design approach that strays from rigorous user testing is not necessarily the cause of the failure. The less you engage with your users, the more risk you introduce into the project, but with that risk comes also the opportunity to innovate and break new ground. It’s up to you whether you choose the steep rocky climb or the safe, well-trodden path, but that the former is riskier and more prone to failure does not make it a worse choice than the latter. This is the path of the vanguard, the path of the designer prepared to suffer failure for the chance to lead. When those in the vanguard fail, their work is seen as self absorbed and out of touch with the marketplace. When they succeed, their work is celebrated as visionary.