AS rumours swirl around the tech world of Yahoo and Microsoft taking yet another tilt at a merger, alliance or other arrangement to compete with Google for online advertising, it's hard not to wonder whether Steve Ballmer and his team are pursuing the wrong partner in the wrong market.
Microsoft's most visible recent plays have been consumer-oriented wins such as Xbox and games, and duds such as MSN and online search, but since 2000 the real growth action has been sales to (and footprint inside) enterprise IT users.
Across all but the top end of the server market Windows has swept aside proprietary Unix competitors such as Solaris and Hewlett-Packard Unix, leaving Linux as the only real rival.
Redmond has also built strong positions in email, databases, small business enterprise resource planning and other small business applications.
Meanwhile, corporate purchases of desktop Windows and Office remain great franchises, even if they are no longer the growth drivers they once were.
This begs an obvious question: should Microsoft be seeking a different way to reinvent itself rather than the right permutation with an obviously reluctant partner that is bleeding talent and falling further behind Google with each day that its myopic and anti-shareholder leaders continue to equivocate?
It's a little more than four years since Microsoft and another enterprise software giant, Germany's SAP, called off secret merger talks because a combination of the two would be too complicated to integrate and manage -- and too powerful to pass anti-trust scrutiny.
During those four years, rumours of renewed interest from Microsoft and/or SAP in a deal have surfaced periodically, but nothing concrete has been confirmed. Nor have those impediments to a deal disappeared.
Still, a lot has changed since 2004 to make a deal like the one mooted back then look a lot more plausible:
* Oracle has used acquisitions to build itself into a potent counterweight to SAP in the market for high-end enterprise planning, process automation and business intelligence apps -- leaving Microsoft, IBM and HP in the dust, and possibly reducing anti-trust concerns.
* As a result, the once-solid relationship between Oracle and SAP has cooled and at times become downright hostile.
* As it has tried to broaden beyond large corporations and governments, SAP has run into mounting competition for smaller sales from software-as-a-service plays such as Salesforce.com and Netsuite (and from Microsoft's Dynamics packaged software, derived from acquisitions of Great Plains and Navision).
* With Gmail and its initial iterations of ad-supported online productivity apps, Google has made it clear it plans to attack Microsoft's corporate apps business.
* Microsoft, which had never spent more than $US1.5 billion on a purchase at the time of the 2003-04 talks with SAP, has shown a willingness to do much bigger deals, such as 2007's $US6 billion purchase of web marketer aQuantive and its recent $US45 billion bid for Yahoo.
* Microsoft also has a lot more experience of integrating acquisitions, as a result of bedding down the deals it did not long before the SAP talks (Visio, Navision, Great Plains) and deals it has done since (TellMe, aQuantive, FAST).
* Anti-trust concerns among US and EU regulators over most packaged software markets have faded because of the rise of web services and Microsoft's efforts to settle cases and play nicely with others.
At a more abstract level, focusing on the enterprise (including enterprise desktops) amid the emergence of web services and software-as-a-service seems a terrific way to sidestep Google's strengths and turn what today is depicted as a simple face-off between packaged licensed software and ad-supported services into a more complex game.
For a start, it is hard to see any but the smallest businesses switching to workplace desktop apps paid for by trailing enticing (and distracting) ads in front of their employees.
(Unless the business itself is the main entity making money when its employees click on them?) Far more subtle models will be needed than in the consumer space.
Of course, there are also plenty of reasons why a Microsoft-SAP deal (which is not the only way Ballmer and his team could make a game-changing, enterprise-oriented play, but certainly the most emphatic and far-reaching option) would involve daunting challenges, just as it did in 2004.
* SAP still costs a bomb: $US63 billion, or half as much again as Microsoft's withdrawn full takeover bid for Yahoo. Moreover, there is no particular sign that its leaders are seeking a sale.
* As the unwinding of DaimlerChrysler reminded everyone, cross-border mergers, particularly between companies with vastly different cultures and histories, are notoriously difficult to make work.
* Neither Microsoft nor SAP has as yet shown much success in the embryonic software-as-a-service space, although both seem to have all the skills and competencies needed to do so if they can face the resulting cannibalisation of their existing business models.
* Both Microsoft and SAP face increasing competition from open source (although up to this point it has posed little direct threat to SAP).
* Putting together two leviathans makes the law of large numbers that much harder to overcome if the objective is flashy growth rates -- although realistically, neither company is really priced as a growth stock any more.
Perhaps the biggest hurdles to a renewed look at SAP (or any other enterprise-oriented alternative to the current blinkered pursuit of Yahoo, Google and online ad revenues) is that it would represent such a reversal of Microsoft's recent rhetoric.
Tilting Microsoft further towards enterprise IT and away from its long dominance of the consumer market, in a sense, also requires a rewiring of the company's DNA which might be beyond Ballmer.
As so many recent retrospectives on Bill Gates have underscored, the lasting greatness and achievement of the company he founded rests on its success in making computing accessible to everyone: in putting a PC on every desktop.
To turn away from that heritage, even in the sense of skewing resources and future priorities towards a more focused set of users who happen to be willing to hand over cash, would not be easy.
It might, however, be necessary if Microsoft is to retain its dominance and market clout as the pursuit of Yahoo moves into its sixth month with no end in sight and no assurance of success even if a deal does get done.
