Angela Eager investigates how companies can best gain operational value from software as a service, both now and in the future.
The software-as-a-service (SaaS) market of today bears little resemblance to that of four or five years ago. Salesforce.com, Salesnet, Upshot - subsequently acquired by Siebel - and RightNow were all offering simple, single-function hosted customer relationship management. Today they have grown into complex multifunction organisms, offering extensible platforms alongside straightforward CRM application services.
Many vendors have been proving that SaaS is not just a CRM issue. NetSuite offers CRM as part of an integrated front to back-office suite and there are vendors offering hosted enterprise resource planning, human resources, storage resource management, sourcing, procurement and collaboration services. Even Google is offering storage-on-demand facilities.
More recently the model has attracted tier-one vendors. SAP has launched a functionally limited CRM-only service; Microsoft has belatedly realised the potential, particularly among the SMB area, and is pumping in the R&D; dollars; and Oracle is evangelising on-demand following its acquisition of Siebel.
As well as supporting Siebel's original native CRM on-demand service, Oracle says it will offer the Siebel 7 platform as a hosted service. Mid-market powerhouse Sage offers the same function-rich CRM application either on-premise or on-demand. Open source providers are also entering the fray: SugarCRM and Compiere are the most well known but a visit to Sourceforge.org reveals numerous alternatives.
SaaS is a serious business. Analyst IDC estimates that 76% of the growth in the CRM market between now and 2008 will come from the on-demand market and that vendor revenues in this sector will grow from $700m to $3,6bn over the same period. Although the overall size of the CRM market is not expected to grow dramatically, a shift in spending patterns is expected as the proportion of budgets spent on SaaS rises at the expense of on-premise software.
IDC notes that 79% of those surveyed have used or are considering using SaaS. Gartner also anticipates major growth, estimating that by 2009, businesses will be spending nearly $1bn on CRM as a service, and that 33% of all small to medium-sized businesses will have opted for a hosted model. Statistics like those are why growth-seeking tier-one vendors have moved into the market.
On-demand is accepted as one of several available deployment models and users see it as a way of extending their IT operations to fit their strategy. A less charitable view is that it addresses some of the problems of traditional CRM by providing a rapidly available, working and user friendly system. SAP admits that its enterprise CRM application did not deliver the quick, cost-effective results customers needed from their CRM system, hence the on-demand service.
One of the recent changes in the market is that the hybrid on-premise/on-demand model is becoming a reality. Although Siebel was the first to lay claim to the concept, its on-premise and on-demand offerings were based on different code bases. Sage followed, offering an on-demand/on-premise CRM application based on the same code base, and now SAP has done the same.
The hybrid model has immediate appeal because it offers choice and the ability to move between modes as required. However, not too much should be read into this as evidence suggests that customers do not change modes. For example, although RightNow offers on-demand and on-premise solutions its business is overwhelmingly on-demand and customers do not switch between modes.
What is important is that where a vendor offers on-premise and on-demand options, they use the same data model so that deployment modes can be mixed and matched across a business, according to business need.
DuPont Chemical is thinking about adopting a hybrid approach, but its decision is due as much to shortcomings of traditional CRM software as enthusiasm for on-demand. An existing SAP user with a successful back-office implementation, when it came to exposing DuPont's back office to the front office, it rejected SAP CRM 4.0 on the grounds that it was heavy, too hard to implement and too expensive.
It had taken on Salesforce.com in various departments to solve immediate problems and is now of one of the early adopters of SAP's on-demand offering, which it plans to use in addition to Salesforce.com. The company perceives potential value from the common data model available across SAP CRM on-demand and on-premise, and possible advantages in running a hybrid deployment in the future.
The service is important but so is the degree of vendor commitment as this indicates whether they see on-demand as an adjunct or a core business component. Although SAP has entered on-demand, it is a tentative entry and its technology and business models are oriented around its on-premise business. Julian Johnson, SAP's VP of CRM for EMEA, says on-demand CRM is one facet of a broad enterprise strategy.
"Large enterprise customers see the benefits of a CRM strategy and are looking at a total enterprise-wide CRM solution. [Some] departments need to act quickly and on-demand allows small departments to quickly take something on. They can use it as an on-ramp to a full CRM solution over time."
Another significant change in the market has been Bill Gates plunging Microsoft into the on-demand world late last year, as revealed in an internal memo where he warned of the disruptive threat from SaaS. Chief technical officer, Ray Ozzie, also emphasised its importance, along with his concerns that Microsoft's services strategy was retarded. He asked each of the Microsoft business units to start mapping out a strategy for developing their own services.
Gates reiterated the SaaS message at the March 2006 Convergence user conference when he said: "Software as a service is a very important trend... We believe in it a lot." However, the company aims to offer software "on a server basis and a service basis", he said. "We do not think there will be a swing from one model to another model. As users get smaller in size, maybe the off-premise gets relatively more attractive. We just want to give people that choice."
For Microsoft, SaaS is not just CRM, it includes everything from the Office suite, to hosted Microsoft Exchange, to next-generation MSN services and the Microsoft Windows Live and Office Live initiatives. CRM 3.0 is available as a hosted version from Microsoft partners and the next version, Titan - due in the first half of 2007 - will sport the badge of a native on-demand application and provide a multitenant architecture.
Microsoft has an unusual view on what constitutes SaaS. Rather than it being a complete application, it views 'mash-ups ' and other types of composite applications as prime SaaS offerings that could be delivered under the Live banner. According to Microsoft Business Solutions corporate VP, Satya Nadella, "Live is not just about hosting. It is also about mash-ups. The real idea of Live is exposing the business logic and there are lots of hybrids, like the recently released RSS generator for CRM, which can be very powerful for things like lead sharing and visibility into the CRM pipeline."
Microsoft has taken a trend and created a massive vision, but it has a poor record of translating vision into reality, as its frequently changing project Green/Dynamics and CRM developments show. The company has decided not to host its applications itself or even establish a partnership - as SAP has done with IBM - but to leave its partners to provide the hosting infrastructure. With the issue of the data centre becoming more important because of its effect on availability and performance, this smacks of indecision or muted commitment in real terms.
The current question is not the validity of the model but the potential value. Anthony Lye, group VP for Oracle-Siebel, believes that the CRM market is characterised by four buying types: deep verticals, simple modular buyers who buy ERP/CRM as a single instance, those who want front to back-office integration, and those who consume components to build custom applications.
Anthony Lye, Oracle-Siebel group VP
On-demand is tied to three of the segments, he argues, with the component buyer being excluded because it does not offer a value proposition.
The on-going debate over whom on-demand suits best - large or complex operations that need extensive customisation capability or more straightforward SMBs - is nowhere near a conclusion. On-demand is still a learning process for vendors as well as customers.
Gartner analyst Michael Maoz says CRM on-demand has demonstrated clear business benefits for organisations with simple-to-moderate deployment requirements but remains unproven for larger, complex environments. "We do not see enterprise businesses going fully online any time soon," says Lye. "[On-demand] needs maturity in the model and business processes."
That helps to explain Oracle's approach. While it offers classic CRM on-demand - on an application specifically built for deployment over the Web - via the Siebel operation, Oracle has also provided its own version of on-demand since 1999, based on its standard software with outsourced maintenance. It is quietly transforming its business, has about 600 developers working on on-demand and believes that more than 50% of its business will come through this mode, However, the company is preoccupied with its huge Fusion middleware and application developments.
Despite recent expansion, on-demand vendors are still essentially single-function providers and this raises important issues. Organisations that plan to make extensive use of the model will probably consume several services. While each may work well by itself, they may not align well, resulting in functionality gaps or overlaps.
Maoz stresses that it is important to identify and document complete CRM requirements to a three-to-five-year horizon to maximise end-user productivity and minimise the cost of ownership.
Sage has the potential to address on-demand cost creep by taking on-demand beyond CRM through its new Sage 1000 integrated ERP/CRM suite. Although on-premise at the moment, there is potential for this newly architected suite to be offered on-demand in the future. SAP is also thinking beyond CRM and considering cost issues and may offer its mid-market all-in-one solution on-demand next year.
Irrespective of the deployment mode, CRM should not be a standalone application. As on-demand deployments become more strategic they need more extensive and wide-spread integration with other enterprise assets. Equally significant are the issues of quality and responsibility. Salesforce.com advocates the concept of the Business Web whereby an organisation can run entirely on the back of services. Its AppExchange platform is a first step towards enabling this goal by facilitating multivendor on-demand services. But again the reality is a mix of multivendor, on-demand and on-premise.
Customers also need to consider the hosting infrastructure. Salesforce.com's recent outages highlighted the need for would-be subscribers to look beyond the vendor and start paying attention to the quality of the data centre behind the operation and the data centre provider. The idea that the subscriber does not have to worry about the application and infrastructure only holds true if the provider can deliver to expectations.
This issue also has an impact where subscribers are considering open source on-demand services. The on-demand aspect eliminates some of the common open source worries such as in-house skills levels and the level of engagement needed by the user organisation because the setup and management are handled by the host organisation. But on the down side, open source on-demand vendors are emerging players with limited resources, which has implications for the quality of the data centre.
As Microsoft's Doug Burgum, senior VP of the business solutions group, says, there is more to hosting than the application: "There is a lot to it: a global platform, scalability, service and partner offerings and billing customers. There are a lot of elements in the infrastructure before you get to offer an application."
Doug Bergum, Microsoft business solutions group senior VP
SaaS has a long way to go to reach technical maturity, but we are only at the base of the learning curve in terms of comprehending how different SaaS is from traditional applications and uncovering its potential value. That translates to significant challenges: managing SaaS deployment to gain operational value now, while exploring new ways of thinking about the form of applications and how they can be developed for future change. However, hosted deployments are not foolproof and if the service is not deployed as part of a strategic business and supporting IT plan, it will fail just as often as ill-planned on-premise enterprise application implementations.