I’m old enough (just turned 50) to have been in the computer business back when something that looked an awful lot like “the cloud” was called “time sharing” or a “service bureau.” The former was IaaS or PaaS and the latter was either a SaaS subscription or true utilization model. In fact, for some time in the early days of the business, companies like IBM sold this kind of model regularly.
So the first thing that occurred to me when the cloud came along was, “Here we go again, another new label on a very old idea.” When I looked at it more closely, I realized that, of course, the public Internet created a new kind of market and connectivity for these services. When I thought about it more, I realized that, of course, the glut of bandwidth was really the enabler in some very important ways, making ubiquitous connectivity virtually free (try buying a T1 or T3 circuit 15 years ago). But it’s still the same model as time sharing or service bureaus, and I’ve yet to see any fact or bit of evidence contravene that view.
Those models had two kinds of buyers, predominantly. The first were time- or simplicity-driven buyers who could be convinced that they would get value more quickly, with less complexity/risk by just consuming software rather than running it themselves. The second were essentially arbitraging computing costs, running their applications where it was the cheapest.
I see the same trend in the cloud. Think about it: Why are companies piling onto salesforce.com? If you know CRM, you know it doesn’t do anything functionally that you couldn’t do 15 years ago in best of breed sales force automation systems. Today, most companies really only want a forecast management system and marketing database, and salesforce.com is really good in those areas. But it isn’t considered differentiating or strategic for most companies; it’s actually kind of generic. It’s also much cheaper than the Siebels of the world were 10 years ago. The motivation wasn’t and isn’t that the cloud enables their sales force to do things they could not do with on-premise software, but rather it’s about cost reduction and simplicity.
Consequently, the motivation for these choices really isn’t about doing things “better.” There is nothing wrong with these drivers, but, in my day-to-day work, I see that many businesses need to do more than reduce IT costs. They also need to deliver better value to the business. Cost arbitrage and reduced time to value are really nothing more than arbitrage if the underlying application/stack remains the same.
So, here’s what I’m left wondering: If your current architecture and platforms aren’t delivering the agility you want on-premise, why would propagating the same kind of technology in the cloud give your company greater agility? You may be able to get lower cost and greater simplicity. However, if we continue to use closed, proprietary-logic applications that turn our business processes into code, undecipherable by mere mortals, how can we expect to become more agile and responsive to business needs?
No, the real opportunity in the cloud is to adopt new architectures by accessing them wholesale, instead of building them yourselves. You may not be able to build your “field of dreams” architecture on-premise, but maybe by selecting a cloud provider who has already done so you could leverage new, more agile platforms, in addition to reducing cost. In this way, the cloud, presents a strategic opportunity for architects that may not come along again for some time. I say take advantage of it while you can.
Glenn Donovan is Regional Director at Cordys. For 25 years, Glenn has helped companies apply complex, new technologies in enterprise computing settings. With roles in companies like Algorithmics, Siebel, SeeBeyond and others, he's helped make the complex practical and achievable. You can follow him on Twitter at @BPMGlenn.