Thursday, January 11, 2007

Common SaaS (Software as a Service) Myths Debunked

In view of Mark Beinoff's interesting perspective to SaaS ala, I thought a post highlighting the common myths associated with SaaS would be in order. I attempt this with a full understanding of the fact that terms like SaaS and Enterprise Software are hazily defined and exist in numerous forms. So take the following with a pinch of salt.

1. SaaS is for low-end, non-mission critical applications
Wrong. As Beinoff himself points out, 15,000 odd folks at Cisco use his service. SaaS is equally potent for business critical applications and in fact many customers are having their major businesses running out of SaaS based applications.

2. SaaS is cheaper than traditional software licensing and homegrown hosting.
Depends on how you perceive and define costs. SaaS has a subscription based model which is an operating expense, while procuring software and hardware is a capital expense, entitling a different tax treatment. In self-hosting and development, there may be lesser recurring cash outflows than you may end up for the SaaS model. In addition being at the mercy of a single vendor like may not be a "cheap" option in the long run. SaaS also introduces "newer" management costs - the cost of vendor management, the cost of not having exactly what you wanted in the first place and the cost of getting the things fixed by the vendor as you would like them to have (you have much more control in a home grown system).

3. SaaS is *the* total IT solution ("End of Software"?)
Well I don't know what is meant by this quote announcing apocalypse. You still need networking software and hardware to be able to use SaaS in the first place. In addition you also need to plug your existing legacy systems to be able to seamlessly integrate with the SaaS model, implying new software development. Do not ever think that you as a customer do not need any additional software - let me quote Beinoff himself:
"We are empowering .. partners and customers to create even more applications with our Apex platform and programming language. That explosion of choice attracts more customers."

4. SaaS is less reliable and more outage prone than self hosted home grown solutions
In most scenarios, barring exceptions, a SaaS vendor can guarantee more reliability and up time that your in house IT department. There are obvious exceptions to this (ex. Google) but in most cases the SaaS vendor will have superior checks and controls as well as more skills to ensure that your systems run uninterrupted. I say this with full knowledge of multiple outages as Beinoff claims 99.9% reliability in spite of them - making a total downtime of 8.76 hours a year. In most cases your IT department will be stretched to achieve this figure.

5. SaaS is a security risk -or- How can I give all my data to that guy?
SaaS vendors are highly concerned about your security concerns. Their whole business hinges on the fact that they can provide a safe and secure computing environment. SaaS is a good choice if you have to deal with tight compliance issues (like SoX). Remember most of us have our credit card data and other personal details on the email servers of Yahoo and Google and still sleep peacefully. Another point to view is how safe you are if you have it in-house - can you be assured of zero failures? And if security is still a concern for you from a business angle (ex. you are CIA) then of course you are better off without SaaS in doing your own thing.

6. SaaS saves me from customizations and implementation blues. I do not need any consulting organization's hourly rates to be up and running.
Maybe not. In most cases a plain vanilla SaaS offering will be of little use to your company. You will probably need some alterations in the base "offering" to be able to derive business benefits out of it. And for this you will need to go back to your friendly neighbourhood IT consultancy. Beinoff himself admits:
"months after we gave customers the ability to change table names and add customer fields, we noticed that they were creating entire new applications: a major online news service was managing their legal department; a chip design company was tracking bugs. These were not things that we had explicitly designed Salesforce to do, but because we made it easy for customers to make these changes, they were taking that power and running with it."
The point remains that you will have to make changes in the base software to meet your business requirements and score over competitors, which beings me to the last one in the list:

7. SaaS offers me huge competitive advantage. I can concentrate on the business and not worry about the rest.
Maybe, maybe now. You need to look at questions like:
- What is my business?
- Who are my competitors? Does my IT infrastructure help me score over them?
- Is IT one of your core differentiators in the marketplace?
- Are your competitors better off in the SaaS model?
SaaS is not a silver bullet. Beinoff claims 27000+ customers and 550,000 subscribers. This translates to an average of approx. 18 users per customer. I am sure many of these will be small time operators. On the other hand, Cisco itself has 15,000 users. It is clear that this model works for some and does not work for others.

See my previous post here. Additions and corrections welcome.

No comments: