Last fall, with an Oracle ERP implementation in full swing, Welch's Foods Inc. began a search for a business intelligence (BI) system to help manage its shipping and transportation operations.
"The entire company was tied up with this gigantic ERP system, so we didn't want to devote any internal resources to doing transportation reporting [and analytics]," said Bill Coyne, director of purchasing and logistics for the Concord, Mass.-based fruit juice producer.
Welch's, which booked $653 million in net sales in 2007, ultimately chose an on-demand BI system from Oco Inc. Based in nearby Waltham, Mass., Oco provides data warehousing and industry-specific BI capabilities in the cloud that can be implemented in just weeks. It was just what Welch's was looking for.
Users access on-demand or Software as a Service (SaaS)-based applications, including on-demand BI, over the Internet as needed. The applications themselves are hosted "in the cloud" on servers usually housed in the vendor's own data center.
"We didn't set out with the idea in mind that we were going to go with a Software as a Service solution," Coyne said. But Oco's capabilities, low cost of ownership and painless on-demand delivery model made the decision an easy one. Welch's began the Oco implementation later that fall, and the system was up and running by mid-January 2008.
With more midmarket companies like Welch's looking for cost-effective analytics and reporting tools, on-demand BI looks poised for significant growth. Major vendors like Business Objects, which recently inked a partnership deal with Oco, are investing more resources in their on-demand BI business to reach this untapped market, companies that stand to benefit from BI capabilities but often lack the IT infrastructure to support an in-house system.
Under terms of the deal, Oco customers like Welch's can purchase additional Business Objects-based functionality, while Business Objects on-demand customers will benefit from Oco's quick implementation process and vertical market expertise, the companies said.
On-demand BI is also significantly quicker to deploy and less expensive than on-site BI, which customers run and manage themselves in their own data center, furthering its appeal to midmarket companies.
A typical on-demand BI system can be implemented in a matter of weeks compared to an in-house system that can take several months to deploy. And a subscription-based pricing model means on-demand customers pay a few hundred dollars a month rather than laying out a few hundreds of thousands of dollars upfront.
On-demand BI capabilities lagging
On-demand BI functionality, however, is generally less sophisticated and less customizable than that of on-site BI applications, analysts agree. That's OK, said Mani Gill, vice president of on-demand BI at Business Objects, because on-demand BI customers have less sophisticated analytics and reporting requirements than large enterprises.
"The [Business Objects] SaaS solutions have a lot of similar types of capabilities, but in no way are they as rich from a feature functionality standpoint as our hosted products are," Gill said. "[But] the requirements of the target market are somewhat simpler."
In fact, Gill said, Business Objects intentionally withheld some functionality from its on-demand offering, which debuted in April 2006 and now boasts 120,000 subscribers, and instead focused on developing an on-demand BI platform upon which partners could build custom applications.
But Business Objects, which was acquired by SAP in January, might have had another reason for making its on-demand BI platform less functionally rich than its on-site applications, according to Neil Chandler, an analyst with Stamford, Conn.-based Gartner Inc. BI vendors with on-demand BI offerings that equal the functionality of their on-site technologies risk cannibalizing their core business, he said.
If customers can get the same functionality in the cheaper, subscription-based on-demand BI model, the thinking goes, why would anyone pay more for the in-house version? "Vendors have to be very careful how they position these [on-demand] products," Chandler said.
That hasn't stopped Oracle, SAS Institute and other major BI vendors from getting into the on-demand BI market. Most of them offer some version of their BI tools on-demand, Chandler said. SAS began offering on-demand BI tools to customers nearly five years ago, while Oracle launched its on-demand BI business in April 2007.
Though some companies are reluctant to let data outside their firewall, on-demand BI is also piggybacking on the growing popularity of Salesforce.com and other SaaS-based business applications. Customers that share critical sales and other corporate data with cloud-based applications are likely to consider on-demand analysis of that data as well, said Gartner analyst Bill Hostmann.
BI vendors are also expanding their on-demand business through partnerships like the one that Business Objects and Oco entered into last week. Chandler expects to see more partnership and reseller agreements among the major BI vendors and smaller on-demand BI companies in the coming years.
As its popularity grows, so will on-demand BI functionality, he said. Still, it will be a long time – five to 10 years or more – before on-demand BI tools reach the same level of sophistication as on-site BI technologies.
Welch's likes the taste of on-demand BI
As for Welch's, Coyne has no complaints about Oco's reporting capabilities. Each night, Oco loads Welch's shipment and transportation data into a hosted data warehouse for analysis via the Web the next morning, he said.
"Essentially, what Oco has in this data warehouse that they've built for us is every single bit of information that pertains to Welch's orders, shipments and freight payments," Coyne said. "It really gives us tremendous insight into our entire distribution operation," which totals between 40,000 and 50,000 shipments per year.
As for possibly tapping into Business Objects' reporting capabilities now available thanks to the Oco partnership, including Crystalreports.com, Coyne said he has no definite plans. "We'll explore it and pursue it if it makes sense for us."
Dig deeper on Business intelligence best practices