Before this week’s Lockheed Martin network breach story intervened, I wrote a column about the strategic dilemma faced by Microsoft from downward trends in both product pricing and new installations for its flagship Windows and Office products. That’s on top of an overall market transition to mobile where Microsoft does not seem to be playing a leading role. What’s Steve Ballmer to do? I think that to thrive Microsoft has to turn itself into a very different company. Fortunately there are archetypes — other companies that have faced similar pressures yet gone on to reach even great corporate success. I think the time is fast coming for Microsoft to emulate Warren Buffett’s Berkshire Hathaway.

When Buffett took command of Berkshire decades ago it was a profitable carpet company. But the writing was on the wall for U. S. carpet manufacturers: labor costs were rising while foreign product prices were declining. Berkshire could have protected its carpet business by accepting lower profit margins, but that would have just been delaying the inevitable. Buffett decided instead to turn Berkshire into a conglomerate with a difference, that difference being its decision to buy or buy into existing successful enterprises, keeping current management in place. Buffett was investing as much in the managers as in their companies.

Buffett used his cash flow to buy a share in the futures of many other companies. Like Peter, he became a fisher of men. Following such a strategy is a very real option for Microsoft.

There’s no lack of money for Redmond to diversify. For more than a decade the company has reliably thrown off at least $1 billion each month in cold cash that could be used for diversification without impacting in any way the day-to-day operations of Microsoft. But what I propose is something more than that — deliberately shaping Microsoft operations to maximize cash flow for external investments. If Microsoft eliminated its many products that don’t make money that would free at least another $1 billion per month and maybe a lot more.

I’m not saying to stop development or even to cut back on the necessities. The strategy I propose would be like that embraced by Hitler in 1943 when he canceled all military R&D that would not bring new weapons on-line within 18 months. While der Fuhrer made many tactical errors this decision was not explicitly among them.

Microsoft should look at its current businesses and commit increased funding to those where it is already first or second in market share and where the segment is showing (or likely to show) healthy growth over time. That means continued investment in a number of specialized server-based back-end products including Skype but reducing expenditures for many clients. This strategy that has served IBM well in recent years.

This means accepting a long decline for Windows and Office just as Berkshire did with carpets. It means abandoning phone handsets, for example, as Microsoft has already bailed on the Zune. xBox would continue, but with an intense concentration on integration with back end services.

Just as an aside here, I don’t expect Microsoft to jump immediately out of the mobile phone business. I didn’t anticipate Microsoft’s deal with Nokia (neither did Microsoft, I’m fairly certain) and that will get a chance to play out. If RIM drops much further I think Ballmer will try to buy the Canadian company and may well succeed. This does not mean, however, that Microsoft can buy its way to a successful mobile phone strategy. The train has already left that station. But Ballmer won’t be able to help himself and it might make the Microsoft phone business into something that could be sold or spun-off 2-3 years from now where today it literally has no value and would simply be abandoned.

What I am proposing is a dramatically smaller Microsoft, probably 25-30 percent of what it is today, though vastly more profitable. Converting corporate mass into energy there’s $70-100 billion to play with and probably another $50+ billion that has been stashed overseas, unable to be brought back to Redmond without a huge tax penalty.

Changing Microsoft into a Berkshire eliminates the downside of having all those profits stuck overseas since they can be used for international acquisitions at what’s effectively a 30 percent discount.

What could Microsoft buy with that $150 billion? A lot of the wrong stuff if the company relies on current management to find the deals. Ballmer is absolutely the wrong man for that job. Microsoft needs at the very least a Chief Investment Officer with a new vision and the clout to shout down Ballmer when it is needed… and it will be needed.

One implication of this strategy is in branding. The Berkshire strategy has emphasized buying established brands and allowing them to operate independently. For Microsoft to be successful with a similar strategy they’ll have to do the same. The Microsoft brand would become secondary over time.

Microsoft shares have been in the crapper for going on 15 years, so maybe finding different brands is a good idea. The company’s current course will not — cannot — change that. Bill Gates has played competition bridge for years as Warren Buffet’s partner. It’s time to expand that relationship. The easiest way to do that, in fact, would be for Microsoft to simply buy Berkshire.

Good idea.