IBM: Clock runs out on massive Siebel implementation, enter SugarCRM

IBM had the largest Siebel implementation in the world and now appears to be diligently unraveling it for SugarCRM and a system called Sales Connect.

Last week at the SugarCon conference, Gary Burnette, vice president of sales transformation at IBM, gave a presentation. In a Q&A session, Burnette reportedly said that Siebel was being tossed at IBM.

The details of Sales Connect are in the presentation embedded below.

JMP Securities analyst Patrick Walravens said in a research note that the IBM move is notable. Walravens said that IBM's move is "yet another indication of the pressure on Oracle's application business."

Connecting the CRM dots, Walravens said that it's likely that IBM decided it wanted a new CRM systems a few years ago. This system had to be interconnected with internal and external social networks, feature analytics and give visibility into the sales pipeline. IBM probably looked at the CRM leaders including Oracle, SAP, Salesforce and SugarCRM. Why Sugar CRM?

Walravens said:

During the Q&A, the IBM executive was asked why IBM decided to go with SugarCRM. He responded that there were three factors: 1) the open source nature of SugarCRM was compelling, 2) IBM liked the user interface and look and feel of SugarCRM and 3) IBM liked the flexibility of an offering that could be run on premise -- providing "flexibility in where and how we run it." IBM obviously has a strong technology expertise and might want to control its application in ways that wouldn't be possible with salesforce.com.

We suspect there might have been some other important considerations such as the ability to use some key IBM technologies like Lotus Notes and the IBM DB2 database - neither of which would be particularly well suited for solutions from Microsoft or Oracle. Another factor we think is relevant is the number of integration points for IBM's CRM deployment.

In other words, IBM needed a CRM system that played well with its own forecasting tools---Cognos, SPSS---and plug into its collaboration software.

Walravens added that IBM's move indicates that "the window is closing on the traditional on-premise, license software model -- and that the window may be closing faster that some vendors expected."

That point is hard to argue. All you have to do is notice how on-premise enterprise software giants have gone cloud crazy with acquisitions. But there is another wrinkle to ponder. Why wouldn't IBM just go cloud?

The short answer is that the economics of software as a service don't quite add up for a company as large as IBM. In a recent interview with IBM CIO Jennette Horan, she said that SaaS economics can be tricky at scale. "SaaS doesn't have a lot of scale for us and what we need to do isn't doable through a SaaS vendor," said Horan.

Specifically, IBM has more than 400,000 employees (it stopped disclosing headcount after its 2010 annual report). That's a lot of employees to pay for by the user, said Horan, who noted it's more cost effective to build a private cloud to deliver applications. And even if the SaaS contract can be worked out there's a customization issue for IBM, which operates in multiple countries with various regulations.