A CIO dilemma: On-premise applications or Software as a Service (SaaS)Submitted by Hans van Nes on Sat, 07/11/2009 - 07:01
It seems that among the CIO community a Berlin wall is created between those who are determined to move their enterprise applications to SaaS-based constructs versus the group who still believes that on-premise applications are still the best way forward. I think that this is not a simple black and white choice but one that depends on a combination of factors. These are the ones I would consider.
The smaller the company, the more likely SaaS pays off. Critical mass in usage is an obvious element here. Think about your purchasing power towards suppliers and initial investments to get your applications installed, customized and stable. You will need a certain returning investment in your application management resources. These are often not or hardly related to the size of your organization and application usage. Twice as much end users are often handled by the same size application management team. Finding and retaining quality resources is always an issue and often required external staffing impacts the operational costs dramatically.
But: for (very) small companies, especially with a flexible home-grown IT-department, the other factors can make on-premise the better and cheaper solution.
It makes a huge difference if the business processes to be supported can largely be managed by commodity solutions instead of fully customized applications. Most organizations will claim that they are different from their peers. The same argument arises when talking about ERP versus purpose build applications. A lot of the SaaS cost advantages disappear when tailor made implementations are required. Don't get fooled by the simplicity suggested for this by the ERP-vendors: the old saying that you have to multiply the cost of ERP software times three to cover customization and implementation, is as much true in a SaaS environment. Going for the standard SaaS version could of course be just the trigger to commoditize your application landscape but will require change management effort.
Dealing with the customization monster and the inherent scope creep, is a factor in the transition from on-premise to SaaS. The more permanent question to answer is if your organization can deal with being “just a subscriber”. Like with outsourcing, you will have to establish a pure demand focused IT-organization which is able to match the push from the business with the limitations of the external supply organization(s). This is far less easy that it initially seems and requires another type of resource than the crew build from the remains of the on-premise times. Although the maturity of the SaaS and outsource partners is growing, automatic contract renewal is certainly not the standard and even in-sourcing tendencies seem to get growing attention.
The relative position of the application in its business lifecycle is a dominant factor in both the sense and timing of transferring. Assuming the underlying technology is not obsolete or a costly dead end road, the most important element to consider is the cost per business transaction. If you have a stable, functional complete and well tuned application and you manage on KPI's like cost per business transaction, a new environment will need time to at least meet these same levels. The question is if your organization is willing to accept that. Objective analysis of the remaining business lifecycle and the potential for lifecycle extension are an absolute ingredient to compare the alternatives.
Cost of transfer
in the above factors some cost issues have already been touched: temporary loss of quality, costs of customization and implementation, and cost of change for the IT-organization. Additionally the cost of de-investing in hard- and software licenses have to be considered. Also the cost of rightsizing the IT-resources (lay-offs of current staff and creating a demand-based structure) can be substantial. Last but not least moving from on-premise to SaaS (or vice versa) is a change project like all other projects and will involve cost of change, change management and risks like any other project.
A balanced look towards the pro's and con's of a move to SaaS based upon all the above mentioned factors is needed to make the right business decision. I'm not against SaaS or any other form of outsourcing, I rather advocate a situational analysis instead of the black and white industry hype.
As always, your comments are welcomed.
back to top more blogs