I've been thinking about a business model based on free software, similar to what Frank Hecker calls "Software Franchising" (Setting Up Shop). And I've been thinking about how this business model could be applied to a specific, relatively large and growing business area: residential networking.

My idea is to build a Linux computer (appliance, gateway, server) and network into every new house. This system could be adapted to existing housing (probably more emphasis on wireless than on structured wiring), apartments, etc. This system would connect to the Internet, and internally could interface to telephone, audio, video, HVAC, security system, lighting controls, and pretty much anything else that the user fancies (and can trust).

Everything I've mentioned exists now, and there are dozens of companies actively trying to foist such systems upon the public. Their progress is being slowed by many factors, including (in no particular order): incompatible technology, high price, poor reliability, lack of training and understanding, and bad software. What all of these problems cry out for is service.

People who think about free software business models gravitate naturally to service, because that's what's left after you free the software. However, if you start from the need to offer a service business, what does free software offer you? For starters: code that is built to need; that adheres to common, non-proprietary standards; that is open to document and understand; that is widely used, broadly tested, and more robust; that costs much less to develop, and next to nothing to universally deploy. So whatever else we need, free software is a necessary part of the solution.

We can identify several other parts to solving the service problem:

• Some activities, such as information dissemination, are very scalable, so a single huge organization has a marked advantage.
• Other activities, such as installation, are one-on-one, not scalable. Here the advantage goes to individual owner-operators, who have the strongest motivation.
• The large organization can lower costs by pooling buying power and playing off competition.
• Small, local operators are more flexible, which can provide a better fit to their customers needs.
• Marketing would be most cost-effective through a single brand name.

When one looks at the industry building new houses, the most striking thing is how diverse it is. In the US, there are no less than 128,000 general contractors building new single-family houses. Those contractors in turn choose from many thousands of subconstractors: electricians, plumbers, masons, sheetrockers, and many other trades. The general contractors and sub-contractors are effectively service businesses, plugged into a market which produces over 1.2M single-family houses per year (averaging $203K each).

Given this industry, the ideal solution nearly jumps out at you: one big organization in the middle, sharing information and software, and acting as an industry-wide coöp; and independent affiliates signed up from the 100,000's of small businesses already serving the industry. (As an added bonus, those same small businesses also work on existing houses and small businesses.) The two levels are bound together by a franchised brand, by investment of affiliates in the central organization, and by open communications. The combination should be both an integrating force in the industry, and an unbeatable value for customers.

So, given that the technology is pretty much ready, that the market is pretty much ripe, and that the business model is obvious, why isn't anyone doing something about this? I think part of the answer is that the businessfolk among us are much too much in love with ideas like high margins, high leverage, get rich quick. While there are ways to squeeze a little extra margin out of this business (e.g., brand management), the basic fact is that one of the great strengths of the free software movement is the willingness to do it for much less.

Another great strength is our affinity for working in the open. I've written up a draft of a business plan for developing and applying free software business models to the residential networking industry, and I've put it up on the web here. As far as I'm concerned, this business plan is free game for anyone who wants to run with it. (Although it should be noted that, in theory, whoever is broadest and most all-inclusive with it will be the most successful.)

One relatively novel concept in the business plan is the proposal to establish a pro-active development fund independent of service revenues. One problem with most well-known free software business models is that they refocus price away from the software, so they tend to provide negligible funding for new development.

A second thing that should be noted is that while Linux advocates joke about world domination, the opportunity to control the home network and to provide the home Internet connection is uniquely strategic. (If that isn't clear, Bill Gates will be unhappy to explain it to you.) What makes this possible is, of course, the trust that comes from free and open systems, and from service professionals who build their business on free and open foundations.

Similar business models could be applied to any service industry where there is a strong technical component. For instance:

• VAR channels, both for generic Linux services and for specific widely distributed vertical markets.
• Local assembled computers (my so-called Linux Craftsmen Guild proposal).
• Web page designers.

Details are left as reader exercises. It will be interesting to see whether others resort to posting open business plans. In my case, this is necessitated by the scale and complexity of the industry, especially in comparison to my own mediocre business skills. It is often said that in business nothing happens until someone sells something. That is certainly true here.