The Cloud

How Low Cost Is That Low-Cost Cloud?

CFOs need to learn how to figure costs per-application to make good decisions about deploying Cloud systems.
Rob LivingstoneSeptember 8, 2011

Today’s Cloud feeding frenzy has been fuelled by heady promises of low costs, almost instant functionality and, ultimately, IT Nirvana.

It’s those putative lower costs, of course, that make most CFOs sit up and pay attention. But if the primary driver behind your Cloud initiatives is to reduce IT costs, then you need to take a second and third look at your financial assumptions.

The Cloud vs. traditional on-premise computing cost argument can be clouded (so to speak) by the way organizations structure and report their IT spend. Those organizations that report IT expenses in the form of the standard chart of accounts, typically broken down into staff costs, depreciation, utility, maintenance, and so on, may not be able to state accurately the actual total cost of a specific application. So if you’re looking to replace one of your on-premise applications with a Cloud equivalent because you think it will be cheaper, then you better be sure that that’s indeed the case.

Drive Business Strategy and Growth

Drive Business Strategy and Growth

Learn how NetSuite Financial Management allows you to quickly and easily model what-if scenarios and generate reports.

Comparing Apples to Oranges
The arrival of Cloud as a technologically viable alternative to on-premise or traditionally-hosted enterprise applications can make for some interesting discussions, but if you are unable to compare costs between applications — typically a per-user per-month per-application calculation — how can you assess whether a particular Cloud offering is low cost compared to its equivalent on-premise system?  

To illustrate this point, here’s a very real world example:

In my most recent role as a CIO in a multinational organization, I transformed the organization’s entire IT cost base to a per-user per-month per-application equivalent by apportioning all fixed and variable costs to specific software applications based on a range of factors such as IT architecture, actual usage patterns, and so on. Then I divided this by the number of users of each system. I did this for all enterprise applications: ERP, finance and leasing systems, web portals, security systems, and so on. When I compared the cost of our Cloud CRM system against all the other systems, I was surprised to discover that its cost was higher than any other on a per-user per-month basis.  This was because a few of the major systems were fully depreciated and I was only paying annual software maintenance. Some other applications were licensed either on a server or maximum concurrent usage basis, which meant that there could be a large number of named users but only a small proportion would be using the systems at any one time.

Given that there was a general perception by the line-of-business managers and some executives that the Cloud CRM system was a far more attractive cost proposition than the on-premise alternative, they thought that a nationwide deployment was logical and inevitable. But using this analysis, I was able to demonstrate that the scale-up cost would, in fact, be anything but cheap, and the nationwide roll-out was limited to specific user categories within the organization.

Unit Cost and Life Expectancy
Another dimension to the discussion of the Cloud’s comparative cost is the total cost of ownership (TCO) over the application’s expected lifespan.

To state the obvious, the TCO exercise for Cloud applications needs to factor in all costs. However, the Cloud may obscure some. For example, there are outgoing system exit costs such as write-off’s associated with the depreciated value of associated IT assets on the balance sheet and early contract termination penalties for existing services. Also, if you need the Cloud application to talk to any other systems, you may also need to subscribe to yet another Cloud application (or hire consultants) to manage data integration, authentication, and so on. 

Once you have developed a full cost profile for the Cloud system, due diligence requires you assess the TCO over the expected life of the system. Large enterprise systems typically are in use for at least three years — many for much longer. By subscribing to Cloud, however, there may be an undue emphasis (at the expense of the full life-cycle TCO) on short-term costs associated with the monthly or annual subscription renewal periods.

At the very least, doing the cost due diligence with some rigor you will let you know with a greater degree of confidence that the Cloud’s promised benefits are realizable.The key message here is not to assume that just because it’s Cloud it’s always going to be cheaper than an on-premise equivalent. If you own IT applications and infrastructure that still have some life in them, your switching costs may be far from trivial. If, however, your existing IT is at the end of its life, the Cloud may indeed be a viable, cost-efficient way to go.

As long as you’re fully informed, and do the appropriate due diligence, you’ll have a fair shot at arriving at the best technological and financial decision for your organization.

Rob Livingstone, an experienced and respected CIO, is the author of the book Navigating through the Cloud. Visit Rob at or email him at [email protected].