Microsoft's cloud push: Too little too late on purpose?

Most of us observers expected Microsoft to move to the cloud now, based on the success of Amazon Web Services and Google. But the apparent pace is to offer developers training wheels, little more. The pricing -- sort of important when the whole goal is about economics and productivity -- remains missing.
Written by Dana Gardner, Contributor

Few topics among economists and business leaders engender the same enthusiasm as productivity. Doing more for less seems the universal balm for individuals, businesses and markets. If we all have growing productivity, well, then everything practically takes care of itself. You'll find few dissenters.

Ask 50 economists, however, how much IT specifically has contributed to productivity surge since the 1970s, and you'll get 50 different answers. They know IT has been good for work efficiency, sure, but just how so and in what dollops? No clue.

Is it better, cheaper software? Is it Moore's Law of escalating power and value from micro-processors? Is it the connected hive of the local area network, or the social fabric of the Internet? The behavior shifts to "always on" data access, or knowledge sharing of ad hoc and geography-free collaboration sessions -- are they behind the productivity boom of the past (apparently closing) bull economic cycle?

Yes, of course, to all. But how and to what degree that these complex causes and effects form the formula for overall productivity is as elusive as predicting the weather from butterfly wing beats. Somewhere in the fog of IT's storm into our consciousness, work habits and business strategies lies the answer. The algorithm of IT's productivity influence over individual people and their societal actions has yet to be written. Too many ghosts. Too many machines.

Nonetheless, productivity -- or rather the expectation of whole new dimensions of productivity -- is what has been behind the hype and strengthening embrace of cloud computing concepts the past two years. In a sense, cloud computing is the unifying theory of IT productivity, which has been indisputably powerful if not complexly disjointed over the past 25 years.

Cloud computing takes many of the essential elements of IT-spurred productivity and serves them up in a whole greater than the sum of the parts. Improved utilization, higher efficiency, better packaging, tuned and refined software, exploitation of network effects, viral effects from social networking, less energy per electron server or packet delivered -- these are just a few of the foundations of cloud computing. What's different is that these variables are working much more harmoniously, with common planning and strategic architectural forethought.

A typical enterprise data center landscape is more a window into the past of IT than the future. The chilled realms of raised floors inefficiently demonstrate how design sprung from unanticipated but compelling paradigm shifts in computing stinks. The barely backwards compatible data center of today eats up larger chuck of money doing less actual improvement in productivity.

Innovation is held hostage by the need to keep the transaction processing monitor properly integrated to the middleware so the n-tier architecture can join data from the right hand to the left hand during a sales call using a mobile supercomputer generating pretty pictures that completely dim after 145 minutes.

We are all well aware of the price for rapid technological change as progeny for helter-skelter IT adaptation and advancement over the past decades. It all probably could not have happened any differently, but it also does not need to continue like this.

Cloud computing entices and seduces because it is, after all, quite different. IT has matured and the requirements of the workload are appreciated sufficiently to architect data centers holistically and effectively. Cloud computing unifies the most up-to-date architectural concepts around data center resources of, for and by productivity. Total cost considerations and the careful association of all of the parts and elements -- working in concert -- these are the high-level requirements of an IT cloud. You can build it right for the workload and allow it to dynamically adjust and accept new workloads. It's more than a just the next big thing. It more than able to drag along all the old stuff too.

When the entire support infrastructure is designed properly, with all the technical and productivity requirements aligned, then IT is transformed. Leverage standards, employ best practices, depend on efficiencies of scale -- and more than incremental change occurs. It does a lot more, more flexibly, for a lot less. Cloud offers a whole new levels of productivity, directly attributed to advanced IT. Services can be assembled based on specific work needs independent of the underlying platforms. Less waste, more haste, all around.

Why then is Microsoft tepid in its march to cloud? Why is it "software plus services," not just services? Why would such productivity improvements that clouds afford -- at a time when economic conditions demand rapid transformational advances -- be dribbled out as Microsoft has done this week at its Professional Developers Conference in Los Angeles? What exactly is Microsoft waiting for?

Most of us observers expected Microsoft to move to the cloud now, based on the success of Amazon Web Services and Google. But the apparent pace is to offer developers training wheels, little more. The pricing -- sort of important when the whole goal is about economics and productivity -- remains missing.

How can architects, CFOs, developers, ISVs, channel partners -- in essence the entire Windows global ecology of participants -- move one step forward without knowing the pricing, both in terms of form, direction and dollars and cents?

My cynical side says that Microsoft wants to accomplish two things with its Azure initiatives. One, to suck the oxygen out of the cloud market (get it, Azure ... no oxygen) and slow the pace of innovation and investment around clouds and cloud ecologies. And two, to make sure the "software plus services" transition comes slower than the market demand might otherwise enjoy. Why swap out software (with a 60 percent margin) for services (with a 15 percent margin) faster than slower?

The answer, of course, is productivity. I have not been sure for many years whether Microsoft is focused on its users' productivity more than at a pace set by, well ... Microsoft. The cloud approach may be different than IT as usual from over the past 20 years, but so far Microsoft's approach to productivity ala cloud seems about the same.

How might this all be different? How might the new, more productive chapter in IT -- of swift yet appropriate adoption of cloud supported IT resources -- get going faster?

Microsoft could spur the engine of adoption on cloud ecology use and advancement by providing stunningly compelling pricing for ISVs and enterprise developers to build and deploy their applications using Azure and platform as a service now. Microsoft would help the makers of applications succeed quickly by making the services easily available to the huge markets that Microsoft is arbiter of -- both business and consumer.

I can even see if Microsoft is choosy and favors its tools, APIs, platforms, data formats, communications protocols, and existing partners. Make a Windows-only cloud, fine, but make it.

Apple with its online store (which favors the Mac world in a big) for developers of iPhone applications and services has shown just how powerful this approach can be. Microsoft could become the best friend of every Visual Studio, PHP and Eclipse developer and business by helping create the best applications for the least total cost, all as a service.

Microsoft could decide and declare what applications it will and won't provide as Azure services itself, allowing a huge market for others to build what's left and sell the services on a per-use basis to users (perhaps driving them to consume other Microsoft services). Deals could be made on applications portability, but optimally the market should pick cloud winners based on value and reach. May the best cloud for both developers and users win -- it would be a huge win.

Redmond could help those applications that provide good value find an audience quickly. Maybe Microsoft could sell or share metadata about users preferences and requirements so the applications are even more likely to succeed. That would include making pathways to the vast Web consumer markets via MSN and all its Web services available to those that build on the Azure platform. Maybe Yahoo finds its way into the mix. Microsoft could offer both advertising-subsidized and pay-per-use models, or combinations of the two for media and entertainment companies, for example, to load their stuff up on the Azure cloud, or build their own Azure clouds. Might compete effectively against Google as a result.

To me these only scratch the surface the the vast and rich ecology of partners and customers that would emerge from an accessible and productivity-priced Microsoft cloud. Done right, myriad specialized value-added business services and consumer services would spring up at multiple abstractions on top of the essential base services that the Microsoft cloud and APIs provide. It would be a very good business for Microsoft, but an even better business growth opportunity for all of the other players. The pie would grow, and productivity could soar. Users would bet better apps and services at low and predictable cost.

There could be a vast and rich community that thrives in the Microsoft cloud's ether. Or there could be a dark Microsoft cloud of, for and by Microsoft applications and services. Redmond could shoot for the moon again, but likely the other clouds will get in the way. Why risk the ecology play for trying to have it all Microsoft's way? That's why time is critical.

Microsoft, at least based on the tepid pace of the Azure roadmap as laid out this week, is more interested in hedging bets and protecting profits than in spurring on productivity and providing economic catalysts to rich new potential ecologies of online, services-driven businesses. Any continued delay to cloud is the giveaway of Microsoft's true intentions.

If Microsoft's own business interests prevent it from realizing the full potential of cloud computing, or make it try and damp down the cloud market generally, then Microsoft is a drag on the economy at a time when that is the last thing that's needed. And yet Microsoft could do cloud better than any other company on Earth.

Microsoft needs to decide whether it really wants to be in the software or services business. Trying to have it both ways, for an indeterminate amount of precocious time, to in effect delay the advancement of serious productivity, seems a terrible waste and a terrible way to affect its community.

The not trivial risk for Microsoft is that in five years it won't be leading in the software or services business any more.

Editorial standards