It’s not so much about a cost shift from large up front costs to an ongoing operational cost and is more about a compounding of items.

On premise software generally includes yearly maintenance and support fees of anywhere from 10% to 20% of cost, expenses associated with operating and maintaing infrastructure and staff for the software, and risk associated with maintaining that software.

SaaS does get rid of the up front cost but also removes operational costs associated with on-premise and always ensures that you’re using up to date software. In addition, the aggregation of operations yields economies unachievable by on premise.

These compounding effects produce a significant net savings rather than an accounting slight of hand.

The concerns you’ve laid out are definitely valid… and usually the first concerns to come up when one starts to analyze the real world utilization of the SaaS model from an end user perspective. If I may, I’d like to address your concerns individually as a precursor to a future article about SaaS’s downsides (and how they may be mitigated.)

1) Security concerns. A very valid concern on the part of any organization - and not surprisingly, also a major hurdle for public/government adoption of SaaS. As you may have surmised by our involvement in the SaaS platform space (Apprenda’s SaaSGrid), my assertion is that SaaS vendors would benefit from getting on board with a trusted platform provider, and SaaS end users would benefit from signing on with vendors who are on board with a trusted platform provider. Why is this? Because any security concerns mitigated by the platform are inherently mitigated in the applications it hosts. A quick example, if a platform provider becomes SAS70 certified, then the aggregate effect is that all applications that are hosted within the scope of the platform inherit this certification. The centralization of these hosting management concerns becomes a value add to the vendor, and a benefit to their end users in turn. Other security risks can be mitigated in much the same way.

2 & 3) Both of your points address the concept of running software in a ‘disconnected’ state. Whether it be due to an inactive internet connection or the lack of an internet connection altogether (i.e. air travel). I agree that the operation of SaaS applications in a disconnected state is a concern that cannot be ignored. In talking with SaaS vendors I am aware of several that tout ‘offline’ versions of their applications. The implementation of these ‘offline’ modes vary between vendors, with the common goal of letting an end user disconnect from the core application and work in a minimally-featured version of the application that is then ‘re-synced’ with the online version of the application at a later time. My advice to someone shopping for a SaaS application is to pay attention to whether or not an application includes an ‘offline’ mode.  Pay particular attention to what is meant by ‘offline’ mode, because the true capabilities of this feature are often surrounded by marketing hyperbole.

4) The cost/benefit analysis of the pay-as-you-grow software model. The consensus at this time seems to be that the SaaS model is of most benefit economically to small/medium size organizations. Larger organizations may not recognize the same ROI effect over the course of, say, ten years, as their centralized user base is tremendous and tends to grow faster. However, it is always important to measure the benefit of fast, reliable, and potentially FREE upgrades and patches. Depending on the number of releases for a particular piece of software over the course of ten years, it can be argued that the end user may save considerably on new features/upgrades/patches in their hosted SaaS software that they would otherwise be paying professional service and installation fees on for the traditional on premise model. This is of course due to the release early/release often mentality of SaaS vendors, and the ease with which SaaS vendors can rollout updates for all customers at once. This whole analysis, of course, should be considered on an individual application basis.  For more on the subject, I suggest reading some of Forrester’s papers - such as Ray Wang’s  “Comparing The ROI Of SaaS Versus On-Premise Using Forrester’s TEI Approach” located here: http://www.forrester.com/Research/Document/Excerpt/0,7211,40227,00.html

Thanks again for your comments and stay tuned for a future article diving deeper into the subject.

- Matt

- Security concerns (company information being held on a server that is outside the company’s network)
- Applications are not available if the connection to the SaaS server is down (If the company’s connection to the outside world is down the employees can at least use their normal desktop applications)
- Employees, especially sales people who are out in the field can’t work if they don’t have a internet connection (what happens if your connection is hideously slow because you have bad 3G reception or there is just dial-up available?)
- Is it going to be cheaper to rent a software than just paying a lump sump?
(Obviously depends on a number of factors as the charging model of the SaaS software, the price for the desktop application, additional cost for network bandwidth, lower cost because roll-outs and updates as much easier, etc)

I guess there are many more. But (with my limited work experience) I would assume that the first point will be one of the major obstacles to the adoption of SaaS.

cu
Martin

You are absolutely right, We’ll make sure to follow up with the downsides in the near future.

Stay tuned!