The subscription model is the core, the very heart of cloud computing’s value proposition.

For CFOs, that model is almost irresistibly seductive, especially contrasted with the traditional on-premise, perpetual-license IT model. Not to mention that the on-premise approach is often associated with long project-lead times, high up-front capital investments, and a dependence upon on a skilled, and therefore expensive, IT staff to implement and maintain applications and infrastructure. In the cloud, maintenance, and the skills and costs necessary to support that maintenance, are (theoretically) shifted to the vendor. Of course, the vendor does not provide that service altruistically. And there’s the rub.

In short, not every cloud has a silver lining.

When Clouds Scale Up
Although the barriers to entry to cloud computing are relatively negligible (especially when compared to writing an enormous check to some giant software provider, hiring a third party to move in to help implement the new system, and then spending months convincing everyone to change how they work), it’s important for organizations to know with some degree of certainty what their system’s boundaries and endpoints will be — in other words, what the enterprise’s future business will look like in terms of the way it will be using the cloud. The influence on the total cost of ownership (TCO) of differing subscription pricing models in some circumstances can render a full cloud scale-up prohibitively expensive. Consider the following:

Many cloud offerings are marketed to customers as if they were a utility; i.e., you pay for what you use in much the same way as you pay for electricity. If you don’t run the washer too often, remember to turn out the lights, and lower your thermostat, your costs go down. But some cloud providers, especially software-as-a-service vendors, charge (essentially) for the right to use the system in a per-user-per-month subscription licensing model. Imagine a user who logs into the system once a month to run a report. Under a per-user subscription, that user costs the enterprise the same as one who is continually on the system, every day of the week. That’s the same as paying your electric company a price that assumes that you run an air conditioner 24/7 even though you don’t own one. That would be crazy, but unless organizations have a clear picture of how they will be using the services they’re licensing, not just now but in the future, they will find themselves in the same position as the homeowner paying the electric company for an air conditioner he doesn’t use.

For cloud to be truly cost-effective, organizations that may be considering scale-ups should study the impact of the various software subscription models carefully. Otherwise, today’s cost savings could easily become tomorrow’s cost constraints.

Here’s a scenario: Your business has deployed a cloud application and, based on its successes so far, you wish to make it available to a much larger audience, including customers, dealers, or even the public at large. But, because you didn’t do a thorough examination of this potential upside usage scenario at the start, you discover (a day late and a penny short) that the additional subscription fees are crippling as your vendor priced the cloud system on a per-user-per-month basis. Your project is stuck, as you can neither easily migrate to another system nor profitably continue the rollout to the broader audience. Depending on who owns this project, someone is in for many a sleepless night.

Your goal should be to avoid this situation. And it’s not that hard: all it requires is a bit of prepurchase planning. (Well, maybe more than a bit, but whatever it takes will be worth it.)

It all comes down to knowing your cloud endpoints and boundaries.

Rob Livingstone, an experienced and respected CIO, is the author of the book Navigating Through the Cloud: A Plain English Guide to Surviving the Risks, Costs and Governance Pitfalls of Cloud Computing. Visit Rob at or e-mail him at

Leave a Reply

Your email address will not be published. Required fields are marked *