(Update)
I have to take issue with Paul Gillin’s approach as he discusses whether SMB’s are better off with SaaS or Open Source Applications. If we equate Open Source to downloadable, on-premise installed software, I have no doubt, and have stated it before that the only good answer is SaaS. But, hold on, a few minutes later we’ll see these two options may not be mutually exclusive for long.

Paul analyzes several criteria:

• cost
• speed of deployment
• customization
• reliability
• data ownership
• vendor viability

These are all issues well-discussed on the web, and although Paul does not explicitly say, my reading is that he also leans towards the SaaS conclusion. The problem is that this criteria-by-criteria approach works well with a typical (mid-size) company where some level of IT expertise is present. Small Business America is very different from the web-savy geeky software startups; the majority are more traditional businesses with no CIO, IT department, in fact often without any IT support whatsoever. While the two main obstacles SMB’s face with any on-premise implementation are cost and (lack of) IT expertise, you can’t just translate the latter into cost – i.e. the cost of hiring full-time IT support. The opportunity cost of Management venturing into IT hiring and project decisions instead of focusing on their primary business makes this an impractical approach, leaving us with only one choice: SaaS.

Another issue not discussed in the article is integration. Open Source or SaaS, getting several packages work together requires IT and business process expertise, which typically means hiring expensive consultants. Therefore, I would go one step further: not only SaaS is the best choice for most SMB’s but they should seek to minimize the number of providers, i.e. the best choice is to use integrated All-In-One solutions.

The current undisputed leader in this field is NetSuite, but as they follow Salesforce.com’s footsteps and move upstream chasing midsize businesses, they leave an opening for up-and-coming challenger 24SevenOffice, which focuses solely on SMB’s, and covers a wider range of business functionality than the incumbent.

This is the situation today. Now, let’s revisit the original question: SaaS or Open Source? A tiny startup named SQLFusion is working on making that question obsolete. The dilemma with Open Source: a lot of good applications are available, but they are written by geeks for geeks… you really have to be quite knowledgeable to download and implement them. Example: at one of the startups I am advising I use SugarCRM over the internet. Starting to use it was a no-brainer, but when I looked at the prerequisites and the process of installing it myself, my head started spinning. No way, this is not for me! Open Source Fusion, which I hear is within days of opening for a limited beta will bridge the gap between availability and usability of Open Source Programs, by offering such apps to be used over the Internet. In true On-Demand fashion, maintenance, upgrades all happen in the background, one can start using the programs without implementing them. So it will no longer be SaaS or Open Source, but SaaS and Open Source.

The first incarnation of Open Source Fusion will provide access to individual applications, still leaving the integration dilemma for SMB’s, but the technology under the hood enables the company to later offer an integration layer between the key applications it serves up.

So the future is Open Source Software as a Service. Hm, here’s an ugly acronym: OSSaaS (?)

Update (3/6). Releated posts:

• SaaSsy
• SaaS for SMBs

Update (5/23): Stefan over at The Small Business Blog discusses the issue; his company, WinWeb is expected to offer Open Source apps as a service soon.