SaaS: An Idea Whose Time Has Finally Come � Again

By Julie Craig

(Back to article)

Industry veterans are familiar with the application service provider (ASP) concept. The idea is that a business software company, or its designated partner, hosts an enterprise software offering for end-user IT organizations. This eliminates the need for IT to plan, deploy, support and modify the software in-house.

While the ASP concept has always seemed like a good idea, it never really took off. Now, we're hearing about the same concept with a new name: software as a service (SaaS), a phenomenon whose poster child is Salesforce.com. Is SaaS just hype, or is something really going on here? And if there is fuel behind these flames, what's different? Why has a model that has languished for a decade suddenly gained steam?

SaaS and the Marketplace

Advocates insist that SaaS's time has come, and after attending the Software as a Service Conference (SaaSCon) in Santa Clara in April, I tend to agree. It is clear an emerging industry is quietly but steadily gaining ground and that SaaS may be on track to change the face of enterprise computing.

Salesforce.com CEO Jim Steele believes the SaaS market will represent 25% of the software delivery marketplace revenue within four years. If this is true, SaaS will have an enormous impact on the way IT does business, and on the fortunes of enterprise software companies.

Before Salesforce.com rolled out its first services in 2001, companies that wanted to keep track of sales leads and pipelines installed customer relationship management (CRM) packages such as Siebel's Sales Force Automation. These products were notoriously difficult to implement, with some estimates putting the percentage of CRM licenses relegated to shelfware as 50-to-75 percent. Why? CRM packages are expensive to buy, they are more expensive to deploy, and can be unwieldy to use.

Industry experts estimate that between 65-and-85 percent of CRM project implementation attempts end in failure. The reasons are numerous and varied but when the costs are totaled, a failed CRM implementation is an expensive proposition. Licensing costs start at six figures, and the total deployment costs can hit seven and eight figures with everything factored in.

With a track record like this, it's not surprising companies are seeking alternatives. These deployments are considered to be both high-cost and high-risk, with a history of failures and cost overruns. It's only natural that a paradigm that promises to mitigate both cost and risk has proven attractive, and that is exactly what is happening with SaaS.

Almost 30,000 companies have already exited the CRM obstacle course in favor of Salesforce.com. Although some say that's just a drop in the bucket—Salesforce.com's revenues are growing at a rate of 50-to-75 percent, per year.

And although the enterprise resource planning (ERP) side of SaaS is not quite as well developed as the CRM side, Dave Duffield, the ERP heavy-hitter who founded PeopleSoft in 1987, is in the process of addressing this shortfall. After Oracle acquired PeopleSoft in January, 2005, Duffield went on to found Workday.com. Duffield's Workday is now in the process of developing an ERP package designed specifically for delivery via the SaaS model.

The track record for ERP implementations approaches that of CRM. Because of the complexity of implementation, many ERP companies actually took a hands-off approach and relegated deployment to IT consulting firms. Although this is certainly one way to get rid of a headache of your own making, thousands of "Big Four" consultants put their kids through college on earnings from ERP implementations; some of which were successful and some of which were not.

Popular wisdom says, "What goes around comes around" and it looks like Workday could have the same impact on the ERP marketplace as Salesforce has on CRM. Workday is already starting to sign clients and Duffield indicates that his platform is on track to be fully competitive with SAP (with the exception of a manufacturing module) by Q2 2008.

Enterprise Software and ASP

There is no doubt SaaS is on track to garner increasing market share, but many insiders are wondering what has changed. The good news is products like Salesforce and Workday are not just ASP re-packaged, but totally new products engineered specifically for SaaS delivery.The major disadvantage of the ASP model was providers typically hosted standard vendor products. Oracle (including the Oracle, PeopleSoft and JD Edwards families of software) and SAP have engineered their solutions for single-customer execution environments; for single-tenancy. These environments include tiered, multi-server architectures with specialized servers for applications, databases, and other activities.

With single-tenancy execution, each customer has his or her own separate execution environment. This means if you need one application server to host Company X, you need a second for Company Y and a third for Company Z. The same applies to database and other servers, because the software was never designed for multiple companies to run applications on a single instance. This architecture impeded ASP's ability to scale and the concept dwindled.

SaaS software is specifically engineered for multi-tenancy. Customer implementations can be insulated from one another and still be stacked on shared computing resources. This is a major cost savings that enables some SaaS vendors to dramatically whittle down delivery costs.

SaaS vendors enjoy other advantages as well. They support only one version of their software and deliver their service on a single platform. Many are making wide use of open source in their products, eliminating a large percentage of licensing fees. They are designing delivery platforms capable of piggybacking multiple products on a single delivery infrastructure. Finally, they are making broad use of service-oriented architecture (SOA), which simplifies software delivery and scaling while providing a very flexible platform for ongoing product development.

Overall, the SaaS model brings with it advantages to consumers that are difficult to ignore. The prospect of predictable levels of service at a predictable cost with minimal risk is an attractive one, especially to war-torn IT organizations tattered from the struggle to deploy and maintain enterprise software packages. Although the industry is still nascent, with many SaaS applications focused at the small to mid-sized (SMB) market, they continue to mature in terms of scalability and functionality.

SaaS also gives SMB subscribers an expertise advantage. Smaller companies typically have a relatively few number of IT personnel who wear many hats. With the complexity of today's IT, it's not possible for a small team to have both breadth and depth.

SaaS provides a way to share expertise among multiple customers while still providing high quality business services. Combined with predictable cost, this combination will be difficult to ignore. The bottom line is SaaS is here to stay, and the industry's first glimpse of "IT as a utility" is very tantalizing indeed.

Julie Craig is a senior analyst with Boulder, Colo.-based Enterprise Management Associates, an industry research firm focused on IT management. Julie can reached at jcraig@enterprisemanagement.com.