Don Norman's view of user research - some comments

I've long been an admirer of Don Norman and his work, which spans both academic research and research within product development organisations (in the past, both Apple and HP). I imagine he is one of the most quoted (and quotable) psychologists, certainly the most quoted in design.

Norman likes to stir up a little controversy, and uses his column in Interactions to rattle a few cages ('My columns are intended to be controversial, to cause people to think, to reconsider ideas they take for granted.'). Recently he's been particularly prickly about the role of research in design.

Some of his comments have made sense, for example his dismissal of over-blown personas or characters to bring to life different, potential users of a product or service. Yes, agreed, probably no need to know what car the character drives or where they have their breakfast in order to develop how they might use a hand held device.

Norman gives neat, cut and dried alternatives of quick, invented personas he has used in product development exercises: the student, the single mother etc. etc. No need for research, he says, to understand your personas. And possibly he's right in the well-trodden field of consumer products and services. But what about innovating in technical areas where you, as a designer, don't know the different characteristics of your users? Well-researched personas help you ensure you cover the ground and, just as important, can help the clients you are designing for understand their audience (it can be surprising how much, as experts in their own business, they can miss about their users' needs). Poorly conceived stereotypes, based on partial knowledge, won't enhance your design work.

More recently Norman has questioned the role of research in design. He suggests that most innovations come from technology development rather than from understanding users, and that since most inventions fail initially, typically because they are ahead of their time, user research has little contribution to make to them. As you read his article you can hear him picking off some of the rhetoric of user-centred design process (it's a small enough world for you to know who his targets are).

Norman claims that user research comes into its own later, once a product has been developed, to refine it for existing and new markets. But I just don't see where the controversy is in this. That is what design is, the application of technology to human contexts. The list of inventions Norman cites that have not arisen from user research (including jet engines, personal computers, the internet, cell phones) are, precisely, inventions. Actually, even though they are technical developments, they are mostly humanly motivated: for faster travel, individual computing, information sharing, mobile communication. You might say the same for more abstract inventions such as alloys with certain performance characteristics: motivated by an industry need that, at root, is human.

At the point inventions become accessible as products or services to people outside the circle of their inventors, they need design to fit them to the needs of their mass users. That's precisely where design, with all its contextual background of user, technology and business understanding comes in to the picture. Bill Moggridge, includes in his book, Designing Interactions, a delightful video of Bill Atkinson, one of the designers of the Apple Lisa (and, later, Macintosh) interface. Atkinson describes how, somewhere at the cusp of invention and innovation, user research helped inform his team's decisions about design development: "I think that the way interfaces get better is by designers watching people trying to use them...and it's very humbling...many times we discovered things...that we would not have discovered without that."

In a later modification of his initial discussion Norman describes translational development: where a researcher spots a problem and goes back to pure science to find a solution, translating that science into a form that will solve the problem. His prototypical case, based on Donald Stokes' book, Pasteur's Quadrant, is Pasteur, who carried out pure, original science in order to resolve a practical problem. Yes, humanly motivated invention. Norman argues for more translational development, from research into practice and from practice into research. He sees it most successfully played out in the development of medical technologies and thinks design researchers can learn from their lead.

By contrast Norman is rather dismissive of what you might call 'Edison's Quadrant', where existing technologies are drawn together to solve applied problems. But I can't see why. I think Edison's quadrant is where innovation consultancies should be (applying technology to human problems, bending technology to human needs); and maybe, with the right problem, venturing into Pasteur's quadrant too.

And in this context, I can't see any question about the role for focused user research. Norman claims, it never reveals unmet needs that designers can go about resolving. And yet, I can't think of a single study I've carried out that hasn't revealed unanticipated findings which, if not met directly through the product itself, need to be resolved through the service surrounding it. I do appreciate Norman's concern that researchers might tend to suggest an over-linear relationship between research findings and design solutions. But I also think there are different kinds of research: research that supports organisations' thinking 'obliquely', keeping them aware of their users and the broad issues they face; that research differs from research that focuses on specific product and service development issues. Maybe dressing up one as the other is the problem. But I wouldn't rule out either kind of research out of hand.

One of the challenges of writing for any audience is (and this is not Norman's fault) the likelihood that they will not read articles in full, but use soundbites as a substitute. (I know this isn't unique to the design community but there is, generally, less of a culture of reading and more of a culture of 'looking' - which would, incidentally, also be valuable in the non-design world.) So, when Norman starts an article with the premise that 'research is essentially useless when it comes to new, innovative breakthroughs', there are many readers who may not get much further than that. In a very thoughtful discussion of a recent Economist/Design Council conference on 'design thinking' (and whether it's helpful beyond design) Kevin McCullagh takes Norman's top line i.e. research is useless in innovation (to be fair to McCullagh who, I know, reads thoroughly and at length, this is a tiny part of a lengthy article and the quote does fit with the opinions expressed by one of the conference speakers). But I do think Norman's position requires some dissection, which is what I've tried to do above.

Just to be clear on this: of course, as a researcher working in design, I would say this. I'm also very much in sympathy with criticisms of the ritual application of research where it really isn't needed. I like this quote from a design manager speaking to Kevin McCullagh about the effectiveness of design process in the pre-recession, boom years of the noughties 'Even turkeys can fly in a tornado'.