Modules and Wholes

Interdisciplinary teams can offer more perspectives on an issue, and more insights for creating a design. Fortunately the culture wars between "creatives" and usability engineers during the dom.com era are mostly forgotten now. But tensions remain below the surface, and interdisciplinary teams often involve more horse trading than a single collective understanding. The reason for this is that true interdisciplinary research involves acknowledging what one doesn't have an answer to.

I was recently reading a book on health research and recognized some of the same kinds of alternative attitudes that affect the design community. To simplify, there are two camps: the "holists", who see health as a difficult-to-articulate but complex interaction of mental, social and physical processes; and the "hard scientists" who believe only in hard data that is unambiguous. The holists deride the hard scientists for narrow-minded "reductionism," while the hard scientists dismiss the holists as given to woolly-minded New Age thinking.

What seems to be happening in at least some health research is the formation of real interdisciplinary research. Whereas previously anthropologists and immunologists both studied disease, they did so independently, without consulting the other's work. Now it is common for social and behavioral scientists to work together with biological scientists to study the interplay of biological and non-biological factors on a certain health issue. The upshot is that the biologists often find that the reality is more complex than previously thought, that strict genetic and biological factors don't explain certain variations. The non-biologists also find the reality more complex as well, that folk wisdom is only partly accurate, or that their intuitions about the behavioral side of health effects cannot easily be generalized the wider population.

In the design world, creatives worry about the reductionism of usability missing the big picture. Usability engineers worry the loosey-goosey decision making of creatives plays roulette with design. What interdisciplinary design can highlight, and attempt to overcome, are the limitation of both the intuitive and data-driven approaches. Intuitives can recognize many patterns in "soft data" that are difficult for hard data methods to find. Hard data can miss nuances because it is too aggregated, or looking at an irrelevant or lagging variable. But intuitive insights can be wrong as well, either over-generalizing, or even becoming a false dogma because they sound like common sense when the reality is counter-intuitive.

Personally, I look forward to interdisciplinary design opening up thinking for all people involved in interaction design. Too much faith is placed in best practice rules and methods, or data collection and decomposition. Too much hope is placed on inspiration as the divine source of good design.

I'm back from a week's stay at a "deluxe" hotel in Sydney, right in the financial district. Very nice place, full of smiling staff and polished marble. The hotel even boasted it was voted one of the top 25 hotels in the Asia-Pacific. Clearly, customer experience is a paramount concern for this hotel.

But for all the concern about customer experience, there seemed to be little concern for user experience. On arriving, I looked around for the standard hotel portfolio binder explaining available services, when they were accessible and what the charges were. Unable to find such a binder, I then noticed the serious looking remote control by the television, and thought "this information must be all online." I turned on the television and started surfing. It seemed a struggle to by-pass the pay-for-view movie selections. I couldn't figure how to go back, when while looking for the hotel information I got sucked into various infomercials touting the hotel chain's other hotels in faraway places. After a seemingly random traversal through unrelated screens and meaningless menu option labels, the only piece of hotel-related information I discovered were stern instructions on what to do in case of a fire. Not exactly where I would expect to look for that information, though happily the hotel marble was not ablaze.

We had to ring the concierge to get the information sought, but it wasn't simple either. There are two reasons I'd rather be a "user" dealing with an interactive device, rather than a "customer" dealing with human. First, the human encounter requires certain nice formalities that can be irrelevant to the task at hand, such as when the well-trained hotel staff solicitously enquire how I'm enjoying my stay, even though I've just arrived and am simply trying to get some information. Second, it can sometimes be difficult to articulate what you want; it is easier to scan for and recognize it. The conversation about the folder/binder thing with hotel information (do these things have a name?) did not quickly result in required information. The staff keep asking what information exactly we wanted, while what we wanted was general information we might want to know once we were aware of it. Our ever-attentive hotel staff attempted to satisfy us by giving as the hotel's corporate newsletter, which didn't have any information relevant to us at all. The confusion was finally resolved when the hotel staff realized that we didn't have the mysterious binder of information in our room as it should be. There was a binder, it had the information what we wanted, it just wasn't in our room.

The incident was hardly traumatic, but it was amusing, considering the enormous stock hotels place in addressing the customer experience. While my Sydney hotel focused on the personal touch, it had a klutzy interactive TV that took several minutes to download the balance on one's room charges.

Hotels can ironically misunderstand to the needs of their patrons on account of their people-orientated systems of delivering services. I don't mean to suggest that patrons don't want personal attention for such matters as getting restaurant recommendations and bookings, or theater tickets. But people, apart from the desperately lonely, don't want all encounters to be personal, and hotels don't recognize this.

Labor economist Robert Reich notes he prefers ATMs to bank tellers. He admits what many feel deep down: "I'd prefer to save my scarce social energies for more important encounters." I think some hotel functions fall in the same category. But while ATMs are famous for being simple to use, hotel IT services lack such distinction. Some hotels have embraced wireless technology, but many, including big name outfits in major global business centers, haven't even grasped the importance of having user-friendly information technology available for patrons. Far too many punish patrons for wanting IT access, viewing it as just another way to gouge customers. Hotel business center fees can make the hotel spa services look like a bargain in comparison. My Sydney hotel charged A$36 an hour for internet access. As far as I'm concerned, that's like charging $36 for a newspaper.

Seems I am reading more and more about "mashed" applications: amalgamations of different applications, wrapped together by a savvy integrator. The concept isn't new, though it has recently become common, and seems posed to explode. Developments in software are allowing easier integration of modules designed by different parties: often parties who would never have imagined their respective children playing together. It can result in marvelous creative fusion, but also poses some unique challenges for the user experience.

Mashed applications bring "hacks" (web APIs, RSS syndication) to the masses. Before recently, integration has been difficult. In a recently completed dissertation, Mika Myller writes: "the challenging usability problem of digital environment of everyday life is that people are forced to act as 'systems integrator'. However, from our point of view the problem is not that people have to integrate products but that they cannot or they have to 'integrate' the products most of the time they are using them because there are not enough possibilities (e.g. open interfaces) for people or third parties easily to compose independent products to systems of systems."

Mashed applications empower individuals by integrating different knowledge, sometimes in ways not imagined. Fantastic. But such a concoction can have a life of its own, and without supervision, cause confusion and disappointment.

Panayiotis Periorellis notes that mash applications, known formally as a "system of systems," can be unstable. Consider the case of the travel website, which brings together hotel, airline and insurance offerings. One can integrate systems from different sources, but the goals of these sources are different, and can potentially change without notice. Something as simple as the length of notice to cancel a reservation can differ. From the users' point of view, they are dealing with a single entity, the travel website, and expect a predictable and uniform experience. But the single entity can be a mirage.

I expect the usability issues arising from systems of systems will become increasingly important in the future. Mashed applications offer a lot, but will need to deliver what they promise.

Looking over the latest CHI Interactions on Offshoring, I ask myself: Is usability about people, or data and specifications? That is the perhaps the central question when looking at how global outsourcing might affect usability over the next five or ten years.

Princeton economist Alan Blinder believes the only jobs immune from offshoring are those where hands-on or face-to-face contact is essential. Many standardized jobs can nearly as easily be done offshore by people following detailed predictable procedures. Onshore jobs will be "in the delivery of services where personal presence is either imperative or highly beneficial. Thus, the U.S. workforce of the future will likely have more divorce lawyers and fewer attorneys who write routine contracts." (see, for example, Will Your Job Survive?)

If one sees usability primarily has the collection and analysis of quantifiable data on user behavior, outsourcing these tasks seems possible, given adequate infrastructure. There are numerous firms selling click stream solutions to track user behavior, and VPN technologies are sure to improve to allow better self-administered user tests. There are even a few companies developing remote elicitation tools to collect data on user wants or mental models, to provide some raw data to shape new designs.

For "mature" user interfaces, where a system of specifications has been defined extensively, offshore designers can design variants without problem. Modularity in UIs is good design practice, and makes it easy for third parties to create new UIs consistent with existing ones.

There are inexorable pressures on user interface design to develop practices that yield more predictability, reusability, and speed to the production of user interfaces. These pressures are driving the creation of in-house corporate, and industry-wide standards. And standards are the lifeblood of the outsourcing industry. If a process can be standardized, it can be outsourced.

Despite the pressures to define standards, there are many, many things about UI design that remain, and will likely remain, messy. Usability professions are like divorce lawyers, and UCD practitioners like psychotherapists soothing the traumatized divorcee. People a want to be happy, and formulaic standardized responses will not offer them the satisfaction they seek to embark on a new, happier life with a new technology.

Telecommunications doesn't seem likely to displace the face-to-face communication needed to understand the why of an issue. Self administered questionnaires and remote discussions are hardly a robust source for insight. Innovation is a counterbalance to standardization. Innovation can be augmented by telecommunications, but face-to-face discussion seems vital.

While I fully expect an offshore aspect to UCD to develop, I also believe that context is too crucial for offshore usability to become a full fledged alternative to onsite usability.