Kieran Levis.com

11 January 2010

This post was to have been about Google’s vaulting ambitions and its supposed similarities to Microsoft, but John Gapper’s latest piece in the FT and Chris Dixon’s blog made me think again. While I agree with almost all of Gapper’s observations on open systems and much of what Dixon says, I have reservations about an abstraction both writers favour – commoditization. And, rather to my surprise, I find myself springing to Microsoft’s defence.

Gapper and Dixon argue that Google’s strategy is either to dominate or to commoditise markets adjacent to its core (news, maps, books and mainstream advertising). According to them, this is what Microsoft did in the 1980s and 90s to PC hardware and other markets. That makes for a nice narrative with clearly identifiable protagonists, and is guaranteed to infuriate Google, which prides itself on being the antithesis of its hated rival. My concern is that this theory ascribes more foresight, intent and control over events to both sets of strategists than they actually possess and conflates too many ideas into a single word. Commoditisation is not a magic spell cast by Machiavellian outsiders, but a complex and mostly uncontrollable process.

A commoditised market is normally taken to mean one with many suppliers, all with similar offerings and none having significant competitive advantage. It is almost identical to the notion of ‘perfect competition’ beloved by some economists, where all suppliers sell pretty much the same products at the same price. (But not at all the same as markets where some goods are offered free in order to generate other revenue streams.) Most markets as they become more and more competitive tend towards something like commoditisation eventually: the innovations and capabilities that were once unique are acquired by others and gradually become commonplace. Only the differentiated or the protected survive.

PC hardware is frequently cited as a commoditized market, because the process happened so conspicuously and so quickly at the end of the 1980s, when the market was still growing explosively. Software was certainly becoming more important than hardware, but this was a general trend, not confined to the PC world. The crucial step in commoditising hardware was IBM’s decision in 1980 to base its PC on open standards, where anybody could supply components. Neither it nor anyone else at the time realised that this would enable competitors to produce near-perfect clones of IBM’s architecture, nor that the remorseless workings of Moore’s Law meant that prices would plummet. In this rapidly evolving market that IBM had done so much to kick-start, its brand and capabilities would soon count for little.

Microsoft was the main beneficiary of a commoditized marketplace, not because it had cunningly planned it, but because it had found the first PC operating system for IBM and shrewdly retained the rights to supply it to others. In the course of the 1980s, DOS and its successor, Windows, became the standard everybody was obliged to use. Gates played his hand brilliantly but his strategy was essentially opportunistic. Once IBM failed to make OS/2 the new standard, Microsoft was sitting on the most lucrative monopoly of the age, and leveraged it to muscle in on applications software markets too. But commoditisation didn’t come into it: Microsoft invested heavily in enhancing its own word-processing and spreadsheet programs, but the decisive factors in ousting the incumbents in these markets were network effects and ruthless tactics. Excel and Word remain profitable near-monopolies, even though Google is now sniffing at its heels with its own versions.

There are grounds for talking of commoditisation in the case of Encarta, the encyclopaedia on a CD-ROM Microsoft brought out in 1993, precipitating the collapse of Encyclopaedia Britannicam, but this was an unintended consequence. Gates’ objective was to create a reason for consumers to buy PCs in large numbers, not to damage other businesses. When Encyclopaedia Britannica rebuffed his approaches he bought the rights to a cheap encyclopaedia then sold only in supermarkets. Encarta however turned out to be good enough for most consumers, especially at $50 or less, and EB’s sales plummeted, along with those of most printed encyclopaedias. As EB’s business model depended on low-income families being suckered into shelling out up to $1,500 for a product that their kids hardly ever used, it is hard to paint Microsoft as the villain of this piece.

The war with Netscape was very different. There can be little doubt that Microsoft’s intention was to eliminate by fair means or foul a rival who at one time seemed to threaten its domination of the desktop. (Raju Vegesna seems to think it did this by giving away its browser free, but hardly any of Netscape’s users paid for theirs either: in so far as it had a clear business model, it hoped to make money from selling server software.) One of Gates’ milder methods was asking AOL, ‘How much do we have to pay you to screw Netscape’. Compaq was threatened with not being allowed to install Windows 95 on its PCs. Microsoft succeeded in marginalising Netscape – at enormous cost – but secured little strategic advantage from making browsers free and Internet Explorer the standard. The public exposure of its methods seriously damaged its brand.

I’ll consider in my next post the extent to which commoditization accurately describes Google’s strategy.