The Idea Factory

There was a nice article in the Globe Magazine that I read tonight about startups in Kendall Square. Seems like there's a great place called the Cambridge Innovation Center (CIC) where Tim Rowe gives entrepreneurs a place to grow their company.

I love this idea simply because I love the idea of encouraging smart people to do smart things in the Boston area. Cambridge, home to MIT, has some very smart people with presumably great ideas. The CIC attempts to keep those people close to home, instead of in Silicon Valley.

I'd like to focus on Conduit Labs, the makers of LoudCrowd, a new type of social videogame centered around dance. At first glance, I must say, Conduit Labs really can't go wrong here: Guitar Hero has paved the way for social videogames entering the mainstream of "cool", and LoudCrowd is probably hoping to pick up where Dance Dance Revolution left off.

Unfortunately, this article exposes a fairly serious weakness in the Conduit team: usability testing. Here's one example:

"Conduit Chief Nabeel Hyatt stands over one of the LoudCrowd testers, Lise Caldara...[who says]
"I like the background. It's awesome," she says. But then she hesitates. "I don't know how to start," she says. "This looks cool. Very confusing, but . . . nice."
...
Hyatt prompts her. "Click for dance," he says."

Conduit Labs [See edit below] If you are in this situation, you have failed. Do not be swayed by the success of your product in these tests, because your tests are useless. You should never be lurking over your users while they test the program, and actually telling the users what to do is so fundamentally against the entire idea of the test that I'm amazed Hyatt didn't realize it.
It seems crystal clear to me that if you have to tell your users what to do, if you're confusing your users as to the entire point of the game- there's your usability test. Go back and start over.

Steve Krug is considered to a god of usability testing, so maybe picking up a copy of his book would do Conduit some good. I haven't read his book, but it looks absolutely great, and I plan to pick up a copy soon. The great Jeff Atwood of Coding Horror agrees about Steve.

In closing, Conduit Labs some people really need to re-examine their usability testing. I think they definitely have an idea worth pursuing, and it would be a shame to see it come out in a half-baked, confusing format. Hyatt is not going to visit each user's house and explain the functionality when the software releases, so it better be clearer than it was when the article was written. If they make their idea clearly presentable and it is as catchy as it seems, they will be successful. Good luck, guys.

EDIT:
Conduit comments that the article was embellished. See below.

PS:
Sorry for the lack of posts lately. I just reformatted and reinstalled (by choice, thankfully), so I've been setting up for the past few days, and that will continue for the rest of the week. I'll do a writeup of what I've discovered when I'm done.