Site icon IT World Canada

SaaS BI

Hosted business intelligence trumps the traditional approach of deploying on-premise tools that are often fraught with high cost, risk and time commitment, according to a San Francisco-based Software-as-a-Service (SaaS) vendor. But one analyst warns of hidden costs to a hosted application.

Especially in tough economic times, constrained IT budgets mean that resources needed to implement an on-premise tool are lacking, and the time to value is ever-critical, said Carlos Mena, an account executive with hosted BI vendor PivotLink.

“BI on the enterprise side can be very expensive,” he said.

The risk factor comes into play when purchased costly user licenses don’t end up reflecting actual utilization, said Mena, because the BI tool is “either too complicated or takes too long to get responses that you are looking for.”

IDC Executive Brief

Enabling Better Decisions Through Unified Access to Information. Download it now. Complimentary with registration.Mena said he has witnessed “various degrees of success” with on-premise tools, and blames limitations on the architecture of traditional BI tools. According to Mena, the typical on-premise tool can lack flexibility to integrate with additional data sources or be limited in terms of what the database can store and report on.

“You’ve invested $500,000, $1 million, maybe $3 million over three years, and you only have a handful of users who are making use of the solution,” said Mena.

He cited a PivotLink customer who used to require six to 10 full-time employees just to manage their on-premise system, for tasks ranging anywhere from querying to reporting to system performance.

But while hosted BI is largely driven by cost and flexibility, Mena acknowledged those aren’t the sole factors when a potential customer debates on-premise versus hosted. More conservative businesses, like those in financial services, may have a leadership culture that prefers to keep data on site for reasons of privacy, but Mena said he finds that as the concept of hosted BI is increasingly better understood, there “is less and less resistance.”

According to Dylan Persaud, research manager of enterprise applications with Toronto-based research firm IDC Canada Ltd., the SaaS model is really an approach that allows customers to defer total cost of ownership of the application, and that there are hidden costs the customer should be aware of.

Although SaaS BI may initially appear less costly than traditional on-premise deployments, Persaud warned that “since BI already runs somewhat in a cloud, to integrate your current systems … may be a bigger task than originally thought.”

By that, Persaud means, that from the outset, there is ample work that must be done with the data that’s to be integrated, in terms of the systems it is sourced from and ultimately how that SaaS data will be consumed by users across the organization.

Eventually, Persaud said, economies of scale should lower the cost of a hosted application given the shared infrastructure and shorter time to value. However, he noted that is yet to be proven with hard numbers. “At this time, there is no real data to support what the eventual cost might be as compared to an in-house application,” Persaud said.