Apr

Apart From Some Doubts About SaaS

Last week, I attended a meeting with a CIO of a large company, and the major theme was about Cloud Computing and SaaS.

This was the most debated subject of last year and now we start to see the first concrete actions. The conversation is very emblematic and main doubts persist about the subject, and I think, it is worth sharing it with you.

The first issue appeared at the meeting which now and then arises in discussions about SaaS, does this model actually reduce the expenses of the company’s software. In my opinion, the SaaS model tends to reduce initial costs, why not demand upfront investments in licenses and infrastructure support as possible acquisitions or expansions accommodate the new hardware system.

Furthermore, SaaS tends to remain more economical over time, since we observed that the traditional model constantly demand a significant upgrade (from version x to version y), and these upgrades often comprise 30 to 40% % of initial cost of purchasing software. There are no such upgrade costs in SaaS.

Another point that favors the SaaS model are the support costs, since the model on-premise must be added to total costs, also costs of the professionals that support the hardware and database.

Another question that appeared at the meeting is, is a SaaS system easier and faster to implement than an equivalent system on-premise.

SaaS has the advantage of no need to spend time on installation, system configuration and components attached to a new cloud services.

But a substantial part of the implementation time of an application system depends on the level of intervention in the processes of the business. At this stage, there is no significant difference between a SaaS and on-premise.

A third question was how to integrate SaaS with legacy systems, which operate in the mode on-premise. This is an important variable. Hardly a corporation of medium to large size will make a total migration to SaaS for one day to another – I even doubt whether you can make this transformation in the foreseeable future.

Therefore, integrating SaaS and on-premise is a fundamental aspect of the SaaS strategy and should be viewed with extreme caution. Integration can be simple, with uploads and downloads to and from the cloud via batch processing or real-time via Web services. This last option can create some performance problems, since the transactions will travel over the Internet.

If the SLA permits any variations in performance caused by external and uncontrollable factors (Internet), fine …

In practice, as more and more concrete experiences with SaaS spread, fears tend to decrease, and the model will spread more quickly. Moreover, IT managers should not remain oblivious to the entry of SaaS systems in enterprises.

The ease with which it engages and operates a system in this model, business executives may be tempted to bypass the IT field and start using the SaaS systems directly. But at some point the system will need to integrate SaaS with legacy systems.

In addition, some SaaS as the salesforce, allow the creation of add-on applications through its own platform (like Force.com, which is a typical example of PaaS). These new applications will require professional developers, and IT must be prepared for this demand.

2011 is going to be hot enough in terms of SaaS. Professionals and managers should be attuned to this motion.