Vendor Lock-In

Software has two unusual characteristics that make it unlike any other merchandised (bought and sold) commodity:

• Big (and endlessly growing) fixed cost to create
• Zero additional (marginal, for you economists) cost to distribute

An argument can be made that Microsoft's OEM business (Windows to Computer Original Equipment Manufacturers) is the most profitable business in human history. OEM's (HP, Dell, etc.) buy media from Microsoft, and then pay Microsoft again every time they resell copied of the media. Even after all the allocations Microsoft can throw at it, you are talking gross margins way over 90% and net margins way over 80%. The US Treasury prints money and isn't that profitable.

The problem with software is that anyone willing to put up an initial stake (say, VC money to cover the costs of development) can play the zero-cost-of-distribution game. So how to keep the software-profit gravy train running? The answer is switching costs - the infamous "lock in."

IBM was the first great player of the lock-in game - they sold proprietary hardware (remarkably-popular mainframes) that required proprietary software to run on them. IBM fought a running battle with the US Justice department for more than 10 years after Amdahl and other companies tried to enter the market with IBM-compatible hardware and software. Justice and IBM's own moves away from leasing killed this golden-egg-laying goose.

Microsoft was the game's second great player. BillG's father is a prominent lawyer, and Microsoft's original OEM agreement with IBM for MS-DOS (a rare non-exclusive agreement that gave MS, not IBM, the capacity to determine what was "IBM-compatible") is a document behind the single greatest transfer-of-wealth in human history, and Microsoft has had a 20-year run on the strength of that single document.

"Open Source" is killing Microsoft, and since Microsoft spends billions per year to develop an operating system that sells for $100 (at OEM), it never occurred to Gates et. al. that anybody would be willing to put in $billions in programmer-hours just to give the product away for free. With Linux, that's just what's happened, and Linux has taken a bite out of Microsoft's growth and profitability.

There is a new dynamic that's killing Microsoft even faster than Open Source did, and for lack of a better term I'll call it "social lock-in." MySpace is HUGE, even though practically anybody could copy and distribute the bits that make up MySpace. What can't be copied is MySpace's community. Ditto, basically, for Google - you can copy the service but still not copy the community. IPod/Itunes is another example.

What this means generally for all software developers is that Social Software is the platform of the future. That which was once called "viral growth" is rapidly becoming an essential component of all leading software.

So how can we take advantage of this? How can we get Tranzax in the hands of more developers and users, and take advantage of community effects?

That'll be the subject of my next post...