Breaking the One-Minute Barrier

Among the many provocative conversations I had at the Web 2.0 Summit yesterday, the most provocative was with podcaster/journo/maniac Steve Gillmor. One topic we wound up discussing was the one-minute barrier: why have most people stopped watching Web videos before they hit the one-minute mark?

Here's what I think:

- Most are bad, and not worth watching for very long.

- Most Web videos are poorly-described. You don't know what you're gonna get until you actually click play.

- It's impossible for a Web site/blog to link you to a precise spot in a long video, so if you don't get quickly to the part you want, you get frustrated and abandon it. (One company trying to solve this precise-linking problem is Motionbox.)

But the biggest reason, simply, is that watching more than a minute of video just doesn't fit very well into most people's established Internet usage.

Think about how you surf: you spend 30 seconds scanning a page like MyYahoo or Variety.com or the NY Times, or wherever you start the day. You click off to a number of articles, and perhaps read a few paragraphs from each one. Unless you are writing a blog entry, or composing an e-mail, or maybe filling out a long form, it is pretty darn unusual to spend two or three or four minutes on a single Web page. (The average time that a user spends on a page of CinemaTech is just about one minute -- and that data surely skews a bit long, because of the way visit duration is measured, and because you and I both know there's not much deep thinkin' here.)

Watching a video longer than a minute doesn't fit in very well -- there's always another browser window open, calling to us, or a new e-mail or instant message arriving.

I don't think the one-minute rule is absolute; the OK GO music videos are about three minutes, the new `Extreme Mentos and Diet Coke' video is the same, and `Evolution of Dance' runs six.

But that leads us to another point... video sites like YouTube only tell us the number of times someone has started playing a video -- not the number of times they've watched it all the way through. I don't expect the info about how many viewers abandon a video mid-way through would be interesting to the general public... but it sure would be valuable to people creating videos. At what point do you start losing viewers? That'd be phenomenally helpful in re-editing a piece of content.

One other issue Steve and I touched on: where will people watch videos longer than one-minute?

My expectation is that as you surf on your PC, you'll tag longer videos that you plan to watch later, and be able to send them to your cell phone, your video iPod, or your television set.

Your thoughts?