Tuesday, February 14, 2006
The Economic Dynamics of Software
Just making my way through McKelvey's Economic Dynamics of Software.
In it, she identifies 3 distinct economic models at play: a) Firm-based: company owns the software and infrastructures, b) Hybrid: companies get commercial benefits from publically available software infrastructures, and c) Network-based: no proprietary control over anything. The three real-life examples used are Microsoft, Netscape and Linux.
An interesting quote from the piece is "What differentiates knowledge intensive products from many other industrial goods is, obviously the relative economic importance of knowledge as compared to physical and distributional aspects.": Software is essentially a codified type of knowledge, but are the distributional aspects completely irrelevant?
In discussing Firm-based models she makes the point that there are clear incentives on the company to control development and use of the product, and there are huge benefits in holding control of a vital component of the software that other software services require. The successful firm needs to continuously innovate and incorporate new features in order to stay ahead of competition.
In the hybrid model, speed to market and brand recognition are stressed as key attributes. The argument is put forward that Netscape was an example of such a company, but under pressure from Microsoft's Internet Explorer, it was eventually subsumed into AOL. Netscape is an example of a product company moving into a services company, a theme which is highlighted by Cusomano.
The network model cites Linux as the example and describes how it is fundamentally uneconomic in nature, but how armies of software developers have willingly lended their support towards its development.
Some interesting linkages are made between software and other types of industries: the sharing of academic information is one, and pharmaceuticals is another. McKelvey says that "software which is openly available on the internet tends to develop in a manner which is even more like public knowledge than like a traditional or standardized product." There is a study by Mowery and Rosenberg (1998) that looks into models that seem to have very strong parallels with software.
There is a paragraph where different distribution techniques (and their costs) are discussed: production and distribution costs can be high (if for instance the software is produced in physical form, CD's etc), or very low if made available over the internet as share-ware.
There is also an interesting point made that the competition in the software market is very different, where companies can compete directly with non-profit making enterprises, and where piracy can directly affect the dynamics of the market.
I'm not sure if choosing Microsoft or Linux were particularly good examples, given the extremely dominant natures of these organisations. Microsoft has revenues greater than the next 42 competitors combined, and as such its degree of control is unprecedented. How about looking at a firm-based model that is successful but not as dominant as Microsoft but is still highly profitable? How about an open source collaboration that has evolved since Linux and has a loyal yet more specialised devotee list? The selection of Netscape seems to indicate that hybrid companies are doomed, but is this really the case?
In it, she identifies 3 distinct economic models at play: a) Firm-based: company owns the software and infrastructures, b) Hybrid: companies get commercial benefits from publically available software infrastructures, and c) Network-based: no proprietary control over anything. The three real-life examples used are Microsoft, Netscape and Linux.
An interesting quote from the piece is "What differentiates knowledge intensive products from many other industrial goods is, obviously the relative economic importance of knowledge as compared to physical and distributional aspects.": Software is essentially a codified type of knowledge, but are the distributional aspects completely irrelevant?
In discussing Firm-based models she makes the point that there are clear incentives on the company to control development and use of the product, and there are huge benefits in holding control of a vital component of the software that other software services require. The successful firm needs to continuously innovate and incorporate new features in order to stay ahead of competition.
In the hybrid model, speed to market and brand recognition are stressed as key attributes. The argument is put forward that Netscape was an example of such a company, but under pressure from Microsoft's Internet Explorer, it was eventually subsumed into AOL. Netscape is an example of a product company moving into a services company, a theme which is highlighted by Cusomano.
The network model cites Linux as the example and describes how it is fundamentally uneconomic in nature, but how armies of software developers have willingly lended their support towards its development.
Some interesting linkages are made between software and other types of industries: the sharing of academic information is one, and pharmaceuticals is another. McKelvey says that "software which is openly available on the internet tends to develop in a manner which is even more like public knowledge than like a traditional or standardized product." There is a study by Mowery and Rosenberg (1998) that looks into models that seem to have very strong parallels with software.
There is a paragraph where different distribution techniques (and their costs) are discussed: production and distribution costs can be high (if for instance the software is produced in physical form, CD's etc), or very low if made available over the internet as share-ware.
There is also an interesting point made that the competition in the software market is very different, where companies can compete directly with non-profit making enterprises, and where piracy can directly affect the dynamics of the market.
I'm not sure if choosing Microsoft or Linux were particularly good examples, given the extremely dominant natures of these organisations. Microsoft has revenues greater than the next 42 competitors combined, and as such its degree of control is unprecedented. How about looking at a firm-based model that is successful but not as dominant as Microsoft but is still highly profitable? How about an open source collaboration that has evolved since Linux and has a loyal yet more specialised devotee list? The selection of Netscape seems to indicate that hybrid companies are doomed, but is this really the case?
Subscribe to RSS Feed