Hosted CRM: A Great Debate or Much Ado About Nothing?

With Siebel's recent jump into the hyped market for hosted customer relationship management (CRM) solutions, the debate on CRM hosting has reached a boiling point. Fundamentally, most organizations want to know whether a hosted CRM solution is tenable for their company.
Written by Steve Bonadio, Contributor

With Siebel's recent jump into the hyped market for hosted customer relationship management (CRM) solutions, the debate on CRM hosting has reached a boiling point. Fundamentally, most organizations want to know whether a hosted CRM solution is tenable for their company. Several critical factors - including application "best-fit," growth and migration, IT maturity, IT constraints, and cost of ownership - must be explored prior to making this decision.

META Trend: During 2003, technical and budgetary challenges will impede upgrades and efforts to consolidate disparate CRM applications. Through 2005, CRM application suites will provide 90% of required ecosystem capabilities. Concurrently, Global 2000 enterprises will gravitate toward CRM functionality from incumbent ERP vendors. By 2006/07, the emergence of services-based component architectures will enable organizations to expose CRM services across the extended enterprise.

In early 2000, we analyzed the market for hosted CRM solutions (a.k.a. CRM application service providers [ASPs]), discussing some of the inherent challenges facing the market. Since that time, the market for “generic” ASPs (e.g., outsourcers that hosted other vendors’ software) has completely imploded, generally coinciding with the end of the dot-com era. Today, we are faced with a different breed of “dedicated” ASP (sometimes called MSP, or managed service provider) that builds and delivers CRM applications to its customers, but in a hosted environment. In more recent research, we discuss some of the myths of small and medium business (SMB) CRM, which is applicable to a debate on hosted CRM because the market for hosted solutions is often inextricably linked to SMBs.

The hosted CRM issue came to the forefront in October 2003, when Siebel Systems (with IBM as the initial business partner) announced a new hosted CRM strategy called OnDemand, followed quickly by an acquisition of hosted CRM provider Upshot. In reaction to these moves, Salesforce.com, an early pioneer in the hosted CRM space, responded with aggressive marketing, including a migration path for Upshot customers. Salesnet and NetSuite quickly followed with their own Upshot migration strategy.

The net effects of this market activity are significant. First, it has clearly elevated the discussion of hosted CRM within the broader CRM space - when a market leader (in this case, Siebel; see Practice 2068) enters a space, it is bound to cause ripples. Second, it has created a schism in the market, causing much confusion about which delivery/deployment model (e.g., on-premises or hosted) is “best,” when in fact the decision about whether to host CRM applications is not necessarily binary. This debate has taken on a decidedly religious fervor, due to the recent market activity. Third, it creates confusion as to which type of end-user organization is appropriate for which type of deployment model. Hosted CRM solutions have traditionally been associated with SMBs, and this is an assumption which must now be questioned, particularly given the entrance of an enterprise CRM player in the market (i.e., Siebel) and Salesforce.com’s recent success at selling larger, multi-hundred-seat deals.

The value proposition of hosted CRM applications is to enable users to get up and running fairly quickly (e.g., a few months instead of several months or even years), with minimal configuration (assuming that configuration capabilities are sufficient for adapting the applications to meet the users’ needs), and lower cost. Pricing typically takes the form of an ongoing, per-user/per-month cost with moderate start-up costs. Current hosted CRM solutions also have some fundamental limitations (e.g., limited ability to customize, extend, and integrate; limited functional footprint and migration path). We note that vendors such as Salesforce.com have recently started to tackle some of these limitations by developing more robust APIs and integrated development environment (IDE) extensions, though writing customizations to a hosted solution does somewhat compromise its “hands off” simplicity.

Despite the promise the hosted CRM space is exhibiting - and putting all issues of hype and marketing aside - the fundamental issue for clients is whether to consider a hosted CRM application. To answer this question, the following factors must be considered.

Application “Best Fit”
Due in large part to the hype from marketing-savvy hosted CRM vendors, end users often find themselves making a decision about deployment models early in their CRM application selection process. With CRM budgets not what they once were (though rebounding), and increased scrutiny surrounding the value of large, complex, and costly on-premises CRM implementations, this trend is not surprising. What is surprising is the fact that many organizations seem to have forgotten their courses in “Application Selection 101” (see ADS Delta 956). Applications - both hosted and on-premises - are purchased to support current and future business and technology requirements. Because it is unrealistic to assume that out-of-the-box (i.e., vanilla) capabilities will meet an organization’s needs, configuration, customization, extension, and integration will be required to varying degrees. We note that more flexible architectures and the emergence of industry-specific CRM products are beginning to alleviate, though certainly not eliminate, the need for extensive, intrusive programmatic change (i.e., customization) to CRM applications.

Further, applications are not stovepipes; they have a place in the broader business and IT portfolio and therefore must be tied to the current application portfolio, enterprise architecture, infrastructure planning, and IT operations. Once these issues (and others discussed subsequently) are adequately addressed, it is appropriate to evaluate deployment alternatives and associated cost issues. Conversely, a leap-of-faith investment into a hosted solution because it requires a minimal up-front capital expenditure might lead to long-term disarray, especially if it does not offer an explicit growth and migration path.

Growth and Migration
In the context of this discussion, growth implies expanding an application investment into new areas of the business or investing in new functions or modules from the same vendor. Migration implies moving from a hosted deployment to one that is implemented on-premises. Clearly, we continue to see on-premises CRM vendors expand their product footprints, and some hosted CRM providers (e.g., NetSuite, Salesforce.com) are doing so as well (e.g., into ERP, order management, and billing). The migration issue is a different matter altogether, and up until Siebel’s recent OnDemand announcements, we have not seen vendors adequately address this issue (we also note that Siebel still needs to formalize its migration strategy).

Migration from a hosted deployment to one that is implemented on premises is critical for organizations that do not want to be constrained by a deployment model and require the option to bring the applications in-house if or when circumstances arise (e.g., more complex needs, deeper integration requirements, IT standards adherence, regulatory issues). We therefore believe that it will be incumbent on hosted CRM vendors to provide organizations with this choice, though it will not be easy given these vendors’ fundamental business models and associated ongoing revenue streams, not to mention potential technology issues. Further, it will be incumbent on “hybrid” vendors (i.e., those that support both hosted and on-premises deployments) to develop migration and change-management methodologies - in addition to tools, utilities, and services to support the physical migration of data, business logic, and rules - to support customers seeking to migrate between deployment models. This migration support is particularly important for organizations that have conducted appropriate due diligence (e.g., requirements definition, IT impact assessment) and decided a hosted CRM solution is a viable short-term alternative, given time-to-benefit issues, but can envision (or are required to because of corporate CRM technology standards) moving to an on-premises solution.

IT Maturity and Constraints
One of the fundamental benefits of a hosted CRM application is its ability to play well within capacity-constrained IT environments; such hosted solutions have garnered the most support and traction within SMBs. Yet, though SMBs are typically faced with numerous constraints (e.g., lack of IT resources and skill sets to implement and manage CRM applications; limited budgets; price sensitivity; IT conservatism and investment priority; risk adversity), the CRM requirements of SMBs may not be fundamentally different from those of larger enterprises. The functional, business process, and integration requirements of SMBs can be just as complex as those of a larger enterprise, particularly if the SMB is a multinational company with numerous applications that must be integrated. Conversely, numerous large, global, distributed enterprises have significant IT resources and mature IT delivery capabilities, but ultimately have rather simplistic CRM needs. The decision to host a CRM application should therefore not be based on the size of the company in question, but rather on its current and future business and technology requirements, aligned with the broader application portfolio, and balanced against the IT organization’s ability to deliver.

Enterprises must also be aware that deploying a hosted CRM application absolutely requires IT involvement in the decision-making and application delivery process. Contrary to the sales approach of some hosted CRM vendors - which entails working directly with line-of-business (LOB) managers (e.g., VP of sales) while bypassing IT involvement throughout the sales and delivery process - the IT organization must be involved in certain activities. First, local infrastructure (e.g., networks, desktop computers) is required and must be managed and optimized, which is out of scope for hosted vendors. Second, as hosted vendors attempt to address limitations - pertaining primarily to application customization, extension, and integration (e.g., application development extensions like Salesforce.com’s sforce) - application development and IT operations will need to be involved (e.g., to build, test, and manage extensions, APIs, and custom scripts). Third, aligning hosted CRM solutions with a broader business and CRM strategy, in addition to aligning them with other IT disciplines (e.g., application delivery, enterprise architecture, infrastructure planning, IT operations) that are responsible for overall IT strategy and execution, requires at least some IT involvement during CRM planning and delivery efforts. LOB managers must be careful to avoid making a hosted CRM decision in a vacuum, given the effects that this decision will have on the rest of the enterprise (and its overall business and IT strategy and processes). The hosted CRM decision, despite initial cost attractiveness, the appearance of immediate or tactical needs that can be addressed, or its being viewed as a panacea for IT disenfranchisement, must involve multiple departments, both within IT and the business. Effective governance - defined as the means by which business and IT standards are created, adherence to standards is monitored, and accountability is articulated and enforced - is the mechanism by which enterprises can ensure that appropriate and synergistic CRM decisions are being made.

Cost of Ownership
From an initial startup cost perspective, there is little doubt that hosted CRM applications are less expensive than on-premises solutions. Our analysis reveals that for a typical midmarket on-premises CRM application, total first-year costs are nearly 50% more than for a typical hosted CRM application. In a budget-constrained, yet improving economic and IT climate, this cost differential is significant. However, a different picture emerges when costs are projected over a few years, because the hosted model is predicated on an ongoing pay-as-you go basis (i.e., per-user/per-month costs), whereas the on-premises model is based on a larger upfront capital investment (e.g., licenses, implementation services, infrastructure) and annual maintenance and support fees. When long-term costs are analyzed, hosted products approach an equivalent total cost of ownership (TCO) to that of on-premises products in approximately three years. After this time, hosted products’ TCO will exceed on-premises TCO (see Figure 1).

Our analysis reveals that in the long-run, hosted solutions will be more expensive than their on-premises counterparts. Clearly, end users must conduct appropriate financial analysis prior to making a purchase decision, based on the intended nature (e.g., short-term vs. long-term, future replacement plans) of the investment. Whether CRM investments are expensed or capitalized is an issue for individual organizations and will be based on their accounting principles, risk tolerance, and adherence to certain guidelines. However, if hosted CRM solutions are a real option, users should keep at least one on-premises solution on their final shortlists so that they can reasonably ascertain short- and long-term cost differentials between the two deployment models.

Business Impact: Although both hosted and on-premises CRM applications can provide significant value, enterprises should not fixate on the delivery model. Rather, aligning CRM application decisions to support current and future business and technology requirements, while respecting constraints, is essential to maximizing the potential for CRM program success.

Bottom Line: Enterprises must not be swayed into choosing a hosted CRM solution based on market hype or perceptions of it being less expensive. Several factors - including application “best fit,” growth, and migration; IT maturity and constraints; and cost of ownership - must be explored prior to making a deployment decision.

META Group originally published this article on 8 August 2003.

(see Exhibit 1.

Editorial standards