Alaska started shopping for an analytics tool last year when top executives caught the bug for more and better dashboards, desktop applications that can crunch the numbers and produce actionable reports on the fly. The airline had been doing some of this already with home-grown applications.
It soon became clear, however, that the do-it-yourself method was no longer adequate. Alaska analysts were pulling lists of valuable data, but it still took them days to produce output that executives could translate into action.
This manual process was not only slow, but inconsistent as well. With no uniform set of business rules for analytics, lists and reports were not comparable over time, making it even more difficult to produce reliable dashboard-quality output.
In addition Alaska, like most large firms, has a disparate set of data stores. In-house the company relies primarily on Informix with some Oracle, SQL Server, and a legacy, mainframe IMF system. Then there is the external Sabre system. Alaska pumps all its flight information to the Sabre booking system in Dallas, where most of the reservation activity takes place. Sabre sends a huge amount of data back to Alaska, where it gets funneled into Alaska’s internal systems.
So in picking an analytics vendor, Alaska wanted something that would work well in a distributed, heterogeneous environment. “We ruled out anything that was ‘cube based,’” says James Archuleta, director of CRM at Alaska. The reason, he explains, was that Alaska had already started building its own star-schema data architecture, and proprietary, vendor data cubes would add an extra layer to the ETL (Extract Transform Load) process required to populate the structure.
Archuleta says Siebel Analytics plugged more easily into Alaska’s database, star-schema architecture than other products tested. He also likes Siebel’s meta data layer where business rules for analytics can be created and enforced, adding the desired discipline to the process.
Siebel has been moving into BI since it acquired nQuire in 2001, but its identification with CRM has made it hard to convince customers it is also a pure-play analytics vendor. The numbers are persuasive. Kurt Schlegel, analyst with Meta Group in Stamford, Conn., says Siebel’s license revenue for analytics went from twenty million in 2001 to over one hundred million in 2004.
The fact that Alaska is not currently a Siebel CRM customer also lends credibility to Siebel’s claim that it is now a serious BI player. Schlegel says Siebel has done a good job in making its packaged analytics independent of the Siebel CRM system.
But the CRM image is still hard to shake. Analysts say most Siebel Analytics are applied to CRM data. This is certainly the case at Alaska where the planned applications are for campaign management and customer intimacy. Archuleta says Alaska currently has no plans to apply Siebel analytics to non-CRM realms such as HR or supply chain.
The cynical view is that Siebel Analytics is a Trojan horse. Once in with BI, the company can make the CRM sale. Siebel’s release of vertical specific analytics software for HR, supply chain, and financial management, however, tend to soften this view.
“Siebel is struggling to position itself as a BI pure-play,” says Keith Gile, analyst with Forrester Research in Cambridge, Mass. “The uptake has been slower that Siebel would have liked – not surprising since analytics is a very crowded space.”
Meanwhile, at Alaska, there is now interest in replacing home-grown CRM applications with a packaged solution. The company is looking at a number of vendors, including, Vignette, Epiphany, RightNow Technologies, and Siebel.
Of course Alaska will want the new CRM system to integrate well with Siebel Analytics. “It is true that Siebel couples best with itself,” says Archuleta. “Funny how they design it that way.”
Siebel Analytics, A Brief History
- Siebel buys analytics vendor nQuire in 0ct. 2001
- First release of Siebel Analytics in Dec. 2001
- Release of Siebel Analytic Verticals for H-R, supply chain, and financial management in October 2004
- Siebel announces analytic revenues of $111.7 million, a 44 percent increase over 2003