I read recently that Rajan Anandan, Google’s managing director for India, says that his country will be a ‘Cloud-first’ market for computing. The companies there, in other words, will go from having very little information technology (as is the case now) directly to embracing Cloud computing without ever going through the intermediary steps of mainframe-, mini-, client-server-, or PC-based computing.
Of course, Google employees all over the world are talking up the Cloud, but I think Anandan’s on to something. Just as many countries in the developing world have largely skipped over land-line telephony and moved straight from having virtually no phones at all to having tons of mobile phones, so too might something similar happen with corporate computing.
Established companies in the developing world don’t have much in common with startups in the West, except that neither type has much legacy IT — a large existing base of on-premise computing and all the people, practices, and philosophies that go along with it. And a lack of legacy means a greater enthusiasm for the Cloud. When I talk to former students who starting and running companies, they find the idea of owning a lot of servers and applications somewhere between distasteful and ridiculous.
As I wrote in my recent HBR article, a few Web-centric companies like Zynga, eBay, and Netflix are also deeply embracing the Cloud. So we see this approach to computing taking off in companies that don’t have much of an installed base of IT, and in a few older companies that perceive a sea change.
Most of the rest of the incumbents are holding back and being very cautious about the Cloud. They think it’s immature and insecure, they don’t perceive its great benefits but clearly see the massive costs of untangling their legacy spaghetti to the point that it’s Cloud-ready, and they have great difficulty justifying walking away from on-premise hardware and software, lots of which is fully depreciated, in order to write monthly checks to Cloud vendors.
The incumbents’ perspective is not ludicrous. In fact, it makes a lot of sense. Which is exactly why it’s potentially so dangerous. American industry went through a remarkably similar transition a century ago, and it didn’t end well at all for established firms.
A hundred years ago, American factories were in the process of converting from steam to electric power. It was a long, slow, uneven process. And it was led by startup companies and new buildings — the older ones just couldn’t justify the switch to themselves intellectually or financially — it wasn’t clear why electricity was so much better, but it was clear how much it would cost to convert an incumbent.
And how did the the incumbents fare as the manufacturing industry transitioned from one power delivery mode to another? By 1935, over 40% of the big industrial trusts formed by 1905 had failed, and 10% more were limping along. Of the incumbents that survived, most became much smaller, with market shares declining by a third, on average, by the 1930s.
Many factors led to this deep shift in the balance of power in manufacturing industries early in the 20th century; electrification was one of them. Will Cloud computing be similarly important? Will it contribute to large competitive shifts not only among IT vendors, but also among consumers of technology?
The only honest answer is that we have no idea, but my gut tells me that the Cloud will be a big enough deal to contribute to competitive shifts and turbulence even in mainline, old-economy industries. Its intelligent adopters will be able to change, respond, and scale up more quickly than rivals relying on on-premise computing. Could adopters will also generally be more efficient, more analytical, and better able to capture and share what they know. In the intensely competitive business environment of tomorrow (and today), these advantages will be telling.
So I think that history is going to repeat itself. Over the next few decades the shift to the Cloud is going to trip up a lot of successful, well-managed firms in a lot of industries. There won’t be one cataclysmic event or killer app that does them in; they’ll just find themselves gradually falling behind their Cloud-enabled rivals over time (and it may take a long time) until they no longer matter. They’ll end, in other words, not with a bang, but a whimper.
Do you agree? Or do you think I’m overestimating the spread and significance of the Cloud? Leave a comment, please, and let us know.