brianstorms

« Previous | Next »

Getting Past Eye Candy: Show The Ugly Demo Too

Warren Buffet has been known over the years to be particularly interested in "ugly" and "boring" companies, that is, ones that are not "hot", "flashy", and getting a lot of attention but are nevertheless well-run, profitable, growing businesses. He once said, "be fearful when others are greedy and greedy when others are fearful." There's a notion, I believe, that's tied to this sort of thinking, that I apply when it comes to software design. Be skeptical when the demo is perfect and full of eye candy or suggests IMAX thinking for a tiny screen. Ask for the ugly use case: all text, no images; big fonts instead of small; black and white instead of color. Try to break the design. Find where it cracks.

Microsoft Live Labs has a new proof-of-concept demo called Pivot which they've written about over at VentureBeat. It's sort of a genetic splicing of Apple's iPhoto with Google's visual timeline search under Google News Archives, where you see frequency counts of results represented as vertical bars in a timeline, so you get to see how many results there were in, say, the 70s versus the 80s versus the 90s, or by year just within the 70s (something I have done thousands of times when researching material for my book). With Pivot, they've added the actual "pages" as building blocks that make up the bars. That's cool. I'm pleased to see stuff like this; it gives me hope. There is good thinking in this article and I've seen good thinking lately elsewhere too regarding new and sometimes radically new interfaces for browsing, curating, searching, and interacting with ever-increasingly vast stores of information. The more proofs-of-concepts, demos, ideas, the better. I can't wait for computing to get past the desktop operating system model and the conventional website model (mobile apps are a start, but iPhone and iPad leave me mostly underwhelmed except when they're used as leisure-time information consumption devices).

Microsoft Pivot

I am a big fan of more visual forms of search. While I think Pivot is compelling, what the article shows is what one might call a "sexy" use case. If I were doing the demo I'm sure I would have picked the same use case: highly visual and flashy color magazine covers and photographs. What's not to like? But whenever I see a demo of a proof-of-concept software app, I feel compelled to ask, okay, show me the ugly demo too. Show me use cases that break your whole design model. Show me something that you'd never show at a conference, or in a product video released on the company website if and when the product shipped. Something "boring" but at the same time the "normal" use case. How how well does Pivot work with, say, a hundred thousand text documents (say, all-courier-font legal documents, or 100 years of memos & letters & monographs stored at a university archive), none of which have images or diagrams, and all, from a bird's eye view, look identical?

The article brings up Minority Report, but that actually nudges my skept-o-meter further into the red zone. The interfaces in Minority Report are pure eye candy -- literally pure visuals. It's basically VR meets Final Cut Pro. That might help one search YouTube, but what about text? How well, I keep wondering, would Pivot work for the "ugly" mundane use cases? I wonder whether the creators of Pivot are thinking beyond visually distinctive objects like photos and magazine covers, and are confident that it can be just as effective with all those legal documents, memos, letters, and monographs. And emails, and a hundred million books, and stuff like that.

One admission: I'm as guilty as anyone of flashy demos and prototypes and mockups that depict an ideal use case, typically gorgeous eye candy but that might not work so well for countless courier-font documents. Still, I try to remember to ask the tough question of how to break a design that "feels" perfect, because if it feels totally right, something's wrong.

Now, if only the Pivot demo ran on a Mac...

Thinking of posting a comment to this blog entry? That's nice. But please note: As of January 19. 2014, I no longer review comments. So they will never appear. So don't bother. If you want to comment on something you read here, go to twitter and write your comment there and just include "@brianstorms" in the tweet. I am no longer moderating the comments here; I just assume everything is spam and Intense Debate seems to have abandoned support for its product.

« Previous | Next »