For small and midsized businesses (SMBs) wading into business applications, the choices have never been more tantalizing.
The exploding open source market offers SMBs a variety of new options for automating books and customer records. At the same time, scores of new vendors host applications outside of a customer's site and deliver service over the Internet, an architecture known as Software as a Service (SaaS).
With so many low-cost options in play, there's little reason for users to subject themselves to the tortuous installation and configuration cycles that are typical of big software projects. But open source vs. SaaS can be a difficult decision, based on your wallet and tolerance for complexity.
There are plenty of options. The SourceForge.net open source software database lists more than 250 active projects in enterprise resource planning and customer relationship management (CRM) alone. Meanwhile, the SaaS Showplace has more than 150 listings across 33 application categories.
Finding an answer to the open source vs. SaaS question depends on which factors are most important to you. Here are some pros and cons of each approach:
Cost. For a small deployment or short-term implementation, there's no question that SaaS has the edge. Fees can run as low as $10 per user per month for some services. SaaS also has little or no overhead, and switching costs are low. ''You're talking Microsoft Office pricing,'' said Robert Bois, research director at AMR Research Inc. in Boston.
However, those fees can add up for large or long-term deployments. The cost of renting two applications from a tier-one SaaS provider can easily exceed $200 per user per month. If you're planning to use several applications or have more than about 20 users, open source may be a cheaper alternative.
In evaluating open source, consider the fully loaded costs. You'll need servers, a secure place to keep them and possibly software licenses. More importantly, you'll need people. Experts recommend you budget for at least a half-person to administer the systems; more if you plan to do much custom programming. Salaries for such people typically start in the $60,000 range. And even then, you'll have a learning curve.
''A capable IT person with deep CRM knowledge has probably done two or three CRM deployments,'' says Greg Gianforte, CEO of Bozeman, Mont.-based RightNow Technologies Inc., which sells both open source and SaaS products. ''We've done 3,000 deployments. That expertise is impossible for an IT organization to accumulate by itself.''
Bottom line: Open source may be cheaper only in the long term.
Speed of deployment. SaaS is a slam dunk here. Many services can get you up and running in a couple of days. This is particularly attractive if you're just testing the waters and may want to switch service providers later. Also, SaaS providers can roll out new releases quickly because all customers upgrade at once.
Some open source providers also offer their products as a service. Two examples are RightNow Technologies and Cupertino, Calif.-based SugarCRM Inc. That gives you the option of starting to use the software as a service and moving to a fully installed version later.
Customization. This has historically been the weak spot for hosted services, but that's changing. Most of the major SaaS companies now offer customization tools, but their features vary widely. Some may also require you to learn a proprietary programming language, which can be a negative when it comes to finding skilled staff.
''Customization is no longer a big issue'' for SaaS, says Jeff Kaplan, managing director of ThinkStrategies Inc. in Wellesley, Mass. He notes that in addition to development tools, many providers have professional services and offshore programming relationships that can keep costs low.
Open source, on the other hand, is infinitely flexible. Most applications support the most popular scripting and programming languages. You may also have a large library of available enhancements from which to choose. SugarForge.org, for example, lists almost 800 code modules that extend the SugarCRM open source software.
Be aware, though, that licensing terms may require you to distribute to the general public any enhancements you make. Check terms carefully, because there are dozens of open source licenses.
Reliability. SaaS market leader Salesforce.com Inc. in San Francisco has taken some hits lately for several service outages, but the reality is that SaaS is pretty reliable. All major vendors use hardened facilities, most with redundant power and frequent backups. Smaller vendors may play a little fast and loose with these details, though, so experts advise you to ask questions. It doesn't hurt to demand a performance guarantee. Seattle-based Entellium Corp., for example, gives rebates if it doesn't hit 99.7% uptime.
Duplicating a highly reliable data center can cost big bucks. If you don't have such an infrastructure in place, SaaS is probably a better bet.
Data ownership. One appeal of open source is that you always have control of your data. SaaS providers offer profuse promises that customer data is never compromised and is professionally backed up and secured. But those promises still ring hollow for some organizations that simply will not put their data in someone else's hands. If owning the data is a big issue, open source is the way to go.
Vendor viability. Experts predict that a coming shakeout in SaaS will doom some vendors, possibly leaving their customers stuck with a dead-end application or without access to their data. ''SaaS providers don't necessarily integrate with each other,'' says ThinkStrategies' Kaplan.
This is always a bigger risk with small providers. The bigger companies are more likely to be acquired and their customers supported or moved to a different platform.
If stability is important to you, then open source is probably a better bet, at least when it comes to the leading packages. ''When you buy software and a company goes out of business, you still have a piece of software," Bois said. But beware: The open source landscape is littered with niche products that never achieved a critical mass and were abandoned by users. The safest bet for any market is to stick with the leaders.
Paul Gillin is a technology writer and consultant and former editor-in-chief of TechTarget. His website is www.gillin.com.