is a blog about design, technology and culture written by Khoi Vinh, and has been more or less continuously published since December 2000 in New York City. Khoi is currently Vice President of User Experience at Wildcard and co-founder of Kidpost. Previously, Khoi was co-founder and CEO of Mixel (acquired by Etsy, Inc.), Design Director of The New York Times Online, and co-founder of the design studio Behavior, LLC. He is the author of “Ordering Disorder: Grid Principles for Web Design,” and was named one of Fast Company’s “fifty most influential designers in America.” Khoi lives in Crown Heights, Brooklyn with his wife and three children. You can reach him through one of the services below.+
Why is the software that ships with digital hardware so frequently bad? When you buy a scanner or a printer, for instance, the software included in the box that allows you to interface with that hardware is, virtually without exception, some of the most poorly designed and difficult to be found anywhere.
I was reminded of that this evening when I spent a fruitless hour trying to reinstall scanning software for a CanoScan N1240U that I’ve had for several years. This software is categorically horrific. Even its most recent versions seem as if Canon is living in a Mac OS 9 world; scanned files cannot be named with spaces, and are restricted to thirty-two characters in length. The interface is hopelessly out of date (even though it was never particularly consistent, even, with Mac OS 9’s look and feel) and difficult to use. What’s more, the software comes in two obscurely named parts: CanoScan Toolbox and Canon ScanGear — can you guess the difference, and intuit which must be installed before the other? Neither could I.
The Hard Cash Is in the Hardware
Granted, this scanner is several years old, but I have an Epson Perfection scanner at the office that’s of more recent vintage whose software is almost equally awkward. In fact, I’ve had similar experiences with all kinds of peripherals, whether scanners or printers or digital cameras (imaging products seem to be the most severely afflicted in this way). And I’ll bet I’m not alone.
Why is this? Well, the answer is pretty clear: these companies are in the hardware business, and absolutely no direct revenue is generated from the software which they are obliged to ship with that hardware. In this market, consumers base most of these hardware purchases on the fit and finish of the object itself, and its specifications as written up in marketing materials and product reviews — and not on how well the software will allow them to use that hardware. In short, these companies’ offerings are not seriously evaluated based on its quality, so they bring little enthusiasm to the task of producing truly good software.
In a way, I find that kind of crazy. It would be silly to discount the importance of hardware features, to be sure, but so much of the user experience of any hardware object is in the software. Good software makes for good hardware.
Most hardware companies recognize this to some small extent, but they miss the point. They will include some marginal innovation in the utility or application that accompanies their product (e.g., optical character recognition, one-touch emailing or photocopying, etc.) as a way of getting a leg up on their competitors. But they do so only half-heartedly; these software features are often poorly implemented with savagely rendered user interfaces. Just as often they are poorly supported in the long-run; any software feature that ships with a hardware product is essentially frozen in time, never to be updated, improved or enhanced aside from only the most necessary bug fixes.
U.I. for Business
Clearly, there’s an opportunity for a manufacturer of these peripherals to establish itself as a provider of fine hardware as well as truly evolved, well-designed software accompaniments. Hewlett-Packard, for instance, would have been well-advised to have purchased a small, talented software design studio rather than scraping the bottom of the barrel with its recent Logoworks acquisition, which intends to provide cheap graphic design services to small businesses. As a hardware manufacturer, does it make more sense to enhance the attractiveness of one’s products through design, or to jump into a penny ante services field?
Improved software doesn’t need to result from proprietary initiatives, either. The Common UNIX Printing System project, among others, has demonstrated that, to some extent, it’s possible for financially disinterested users to improve the reliability of third-party hardware in spite of lack of manufacturer support. Wouldn’t a million dollars invested in open source software design yield more for Canon than pursuing a go-it-alone approach in which they’re continually turning out new versions of and applying half-hearted fixes to incredibly bad software? I guess they just don’t see it that way, at least not right now.+