Delusions of openness

Howard Stringer of Sony claims a deathbed conversion to open standards. In his interview with Nikkei Electronics Asia last month, he was quoted as saying:

Customers will refuse to accept it unless the technology is open. Youth in particular really dislikes closed technologies, closed systems and the like. …

Sony hasn't taken open technology very seriously in the past. Its CONNECT music download service was a failure. It was based on OpenMG, a proprietary digital rights management (DRM) technology. At the time, we thought we would make more money that way than with open technology, because we could manage the customers and their downloads.

This approach, however, created a problem: customers couldn't download music from any Websites except those that contracted with Sony. If we had gone with open technology from the start, I think we probably would have beaten Apple Inc of the US.

There was a time when it made sense to divide the market with closed technology, and monopolize a divided market, but that's just not an effective strategy any more. In the Internet universe, there are millions of stars - millions of options that have been created through open technology.

I would agree that Sony was the epitome of a closed technology company, but otherwise the rest of the answer (including the factually inaccurate claims of Apple DRM) was just plain wrong. (Intentionally misleading or merely confused, I can’t tell).

(Let me suspend for now the whole question of what is an “open” standard — the subject of my 2006 book chapter in the Greenstein and Stango book. Suffice it to say some standards are more open than others, but few are fully open or fully closed.)

From where I sit, the higher water mark for open standards was about 15 years ago, when the dominance of open standards was clearcut. We had IETF standards for TCP/IP, SMTP and FTP (with W3C standards for the WWW) that were as open as any standard ever: available free on the Internet for all to use, royalty-free. We had 802.3 (with 10Base5, 10Base2 and 10Base-T) from the IEEE.

Today, the young people Stringer mentions spend hundreds of dollars to buy content for proprietary videogame platforms from Sony and others. They communicate over proprietary VoIP protocols from an eBay subsidiary on a PC with a proprietary OS (either one) running on a proprietary microprocessor.

If they should happen to have bought a Sony digital camera or camcorder, they will be used Sony’s proprietary (and royalty bearing) Memory Stick flash memory card and soon will be recording video in its proprietary AVCHD format.

Now they could buy a Blu-Ray player from any number of players, but a fraction of every player (and disc) purchase goes to Sony and it partners in the patent cartel. Worrying about the June 12 (claimed) digital transition, they then buy an HDTV with an estimated $23 going to patent holders.

Similarly, if they have a cellphone, it doesn’t matter whether it’s a GSM or WCDMA phone (using standards from ETSI/3GPP) or a cdmaOne/cdma2000 phone (using standards from Qualcomm/3GPP2) — with a 5+% royalty (before cross-licenses) going to IP holders.

In their 1999 book Information Rules, Carl Shapiro and Hal Varian noted that users will prefer open standards but buy proprietary ones if sellers make it worth their while (think of a “teaser” rate for lockin costs). Firms can lock in customers if they are careful not to get too greedy. Varian is now putting these ideas into practice as chief economist of Google.

So Stringer’s claims to the contrary, we are passing through an era of semi-open standards when firms prefer to support open standards over their rival’s proprietary standards, but much prefer to establish their own proprietary standards (think Memory Stick.) We’ve had semi-open standards for 20 years — think OpenVMS — and except for an occasional open source platform, I think semi-open will remain the norm. I’d be glad to fly to Tokyo to give a tutorial on the subject if anyone’s listening.