Let’s start with expectations from the real world. If I spend more on a car, I get more. Examples abound: I don’t have to insert a key to start my high-priced beauty; the car senses the key and I press a button to start the engine. Temperature is automatically adjusted to preferences set long ago, and seats mold to my body at the press of a button. Budget cars have none of these niceties. The Ozer Paradox is unique to the high-technology space and refers to the paradoxical reality that as technical products get more expensive, they become less polished and harder to use. While not universal, this dynamic is pervasive enough to deserve recognition and hopefully resolution.
Similarly, if you spring for first-class airfare, you speed through a separate line at security, board first, and drink single malt scotch in a cushy seat while the sweaty rabble slowly board in back. Spend more (or get a lucky upgrade), and you get more.
With technology products, the reverse is true. I first noticed this years ago while reviewing analog capture cards for PC Magazine. Consumer cards that sold for $300 had installation wizards, printed quick-start guides, and polished interfaces. Cards that cost $3,000 came in a plain white box with a floppy disk with drivers you installed manually — no wizards, no manuals, no interface.
In 2012, I was comparing enterprise encoders with consumer encoders. One enterprise encoder was so rough that you couldn’t click a button to encode a disk-based file; you had to type in a URI (uniform resource identifier). There was no way to check FTP or S3 credentials without starting the encode, leading to multiple encoding errors. Other similar rough edges made the $5,000 enterprise-class encoder infinitely harder to use than even a free product such as HandBrake.
I just finished testing multiple webcast and live streaming products, some of which cost $1,200 per event. I also tested YouTube Live. Guess which interface was more polished? With one webcaster, you couldn’t preview the video in a player other than the viewing page (which didn’t become live until 15 minutes before the event) without asking for this feature in advance. That’s like shaving, showering, and dressing for the biggest interview of your life without being able to check yourself in a mirror. You literally had to call an engineer and ask him or her to check audio and video quality.
With another service, you couldn’t preview without going live; its suggested strategy was to embed the player in a nonpublic page for preview and insert the embed code into the actual viewing page just before the event. Pretty smooth, eh? That same service wouldn’t provide the server credentials inserted into your streaming encoder until you were 30 minutes from going live, preventing you from testing these links the day before. YouTube Live hands over server credentials when you create the event, which can be days in advance, and allows full preview in the control room before going live.
When presented with this paradox, software producers respond that the target user for the expensive product is more technical and so can handle a more complicated interface. That may be true if you’re producing routers, but there are no university-trained compression professionals. Instead, compression-related products and services are configured by everyone from the IT guy to the video gal, with web professionals and broadcast engineers in between. We need simple.
Software producers also argue that more expensive products have more features, thus the more complicated interface. Perhaps that’s true, but it’s still no excuse not to make basic operations such as preview and obtaining server credentials as simple and flexible as possible.
Understanding YouTube’s motivation is easy. YouTube Live is a free service; if it doesn’t make the service brain-dead simple, users will scream and it won’t get used. So it clearly made every effort to identify the common pain points and to make its operation as smooth as possible. It’s simply paradoxical that companies charging good coin for their products and services don’t make the same effort.