All this stuff worked for a long time, but eventually the company was unable to sustain its success. I do not know the exact reason, but I suspect one important issue is scaling–as a company grows, an idea has to have much bigger impact in order to have the same percentage impact on revenues and profits. Because companies are ultimately pyramidal, the filitering and administrative capabilitiy at the top is fixed and can only process so many initiatives at a time. (Delegating the project approval process farther down the hierarchy works up to a point, but top managers are held responsible for the firm”s major investment choices and so find it difficult to go completely hands-off.) Hence, there is pressure for each initiative to show bigger sales and profit impact than previous initiatives, so as to economize on management’s scarce attention, even as good ideas in traditional areas are getting mined out. It starts to seem like “innovation is getting harder.”

This syndrome appears to have hit firms as diverse as P&G, DuPont, and now, possibly, Microsoft, all of which built their businesses by launching lots of new products. (One wonders whether Wal-Mart has an encore following its astonishing success in moving into groceries–Wal-Mart is now so big that the only new areas they can move into that would have a noticeable impact on their sales and profits are things like auto sales (a suggestiion a colleague made to me recently.))

Of course, very large firms can specialize in attempting innovations which require very large up-front expenditures and large-scale coordination in marketing and distirbution. Ideas suitable for this come along much less frequently, however, than the typical sort of product innovation these firms were built on.

The problem Microsoft has isn’t resources but organization. It has never been particularly good at innovation within. It has historically been good at recognizing innovation without and purchasing it (with said capital). Almost no major software sold by Microsoft had its origins at Microsoft. It was either developed in a company that was bought, or the product was bought, or the product was copied from a competitor and monopoly position used to sell the product until it could be revised enough times to become a legitimate competitor.

However, it remains to be seen which is the better strategy for a company of Microsoft’s size and goals. Apple and Sun produce lots of cool technology in-house, but they do not have the size and shareholder return that Microsoft does (though Apple is approaching it on a per-share basis).

Not having worked at Microsoft, I can’t say what their organization limitation is, but I would guess it has to do with marketing and project management decisions which favor features which test well for vocal market segments but fails to reward features that do not build upon a known process and require a leap of imagination to grasp how much they will be accepted once in the hands of the public.

That being said, it’s hard to argue with their bottom line. Perhaps keeping investors happy as they approach the end of their market share expansion for their base products is more valuable to them than plowing the money back into research which is already not returning much of a bang for the buck.

“Bell labs concludes transistor technology has gone as far as it can. Sony buys rights. But I’m sure Sony was making a mistake, 40 years from now, Bell labs will still be on top due to this judicious move, and Sony will have just wasted their money.”

Sound familiar?

To me this is a classic demonstration that capitalism and free-trade are not the same economic realities. Capitalists horde capital as a game plan (funny enough, even when they give it away) Free-tradists look to producing more.

Capital is only grease. Respect the capitalists (otherwise they tend to take their ball and go home), but don’t honor them. They are needed, the way garbage collection is, but they do not innovate, nor produce, and they have a tendency to fear the free market.