How To Build A Business Around SaaS

Infrastructure Requirements
VARs on the sidelines may not be aware of the up-front costs involved in preparing their business for delivering SaaS solutions, which include acquiring talent and building a scalable network infrastructure that's capable of handling heavy traffic levels.

"You're basically hosting a data center for your clients, and the reality is, you will have some infrastructure costs," said Eric Berridge, co-founder and principal at Bluewolf Inc., a New York-based ISV and integrator that partners with Salesforce.com. "You can scale, as you bring clients on board, to be more efficient, but you don't need to build to support 100,000 users up front."

Security solution provider Vigilar Inc., last June unveiled Atlas, its inaugural SaaS offering, which handles asset and license management, technical support, log management, authentication management and systems maintenance functions.

To support Atlas, Vigilar set up a 24x7 security operations center and then built out its back-end infrastructure, an effort that included setting up co-located sites in geographically dispersed data centers, said Ryan English, vice president of products and services at the Atlanta-based company.

Vigilar also assembled an internal software development team to build an application on top of its existing ticketing system that allows customers to track their status online, which helps the firm provide better customer service, English said.

Yet, building a SaaS business involves more than just hiring a bunch of technical wizards, Berridge said.

"You have to bring in people that are multitaskers and who can manage multiple streams at once and that think in a very iterative fashion. MBAs don't necessarily thrive in this sort of model, and I've also hired many senior level project managers that didn't make it," Berridge said.

Setting Proper Expectations
VARs getting their feet wet in SaaS should educate their customers as to the realities of the technology, which includes breaking down customers' preconceived notions about SaaS and making sure they're aware of the multifaceted changes they need to promote to make their businesses SaaS-ready.

"Many customers assume that because they're going the SaaS route, they'll be able to leverage the application sooner," said Berridge. "...The challenge is to temper that expectation."

One risk that solution providers take with SaaS is that clients will demand more flexibility, scalability and security in a SaaS model than they would have in the client-server model.

Service outages are another reality of SaaS that newcomers need to account for, and cover, in clearly worded policies. Vigilar's procedures for dealing with service outages are based on proactively communicating with customers when an outage occurs, to ensure that the trust they've built up with clients isn't damaged, said English.

"If our user interface goes down, we notify customers immediately," said English. "While no one wants to see an outage happen, people do want to be notified when it does."

Software upgrades have always been a primary cause of service outages, and customers are asking more questions about whether their data will be protected if an outage occurs, English noted. SaaS providers should always inform customers about upgrade schedules, because this helps minimize the impact of service outages, he added.

Drawing up service level agreements (SLAs) that are fair and realistic helps strengthen the customers' trust in SaaS providers, English said. "We're extremely realistic in our SLA. Some companies go too far, and it becomes more of a marketing tool, but it's pretty dangerous when you're setting your uptime claims too high," he said.

With SaaS, VARs that used to conduct on-site software deployments can deliver services without stepping onto the client's premises, and Workopia's Lee said that makes the issue of setting proper expectations even more important. "Some customers expect because everything is online, they can click a button and their entire business gets automated. But that's unrealistic," he said.

Software is still software, even when it's hosted in the cloud, said Lee, and SaaS deployments require a combination of training and resources between the VAR and the client.

Next: Potential Pitfalls