Mychelle Mollot: The key thing we say today in regards to ERP is people are now moving away from focusing on transaction systems which were about automation to focusing on optimizing and that’s really where the business analytics capabilities come in. Its about the difference between automating something to make it work to ‘I’m going to make things work better and improve my outcomes and drive a better future for my company.’
CWC: Recent acquisitions by IBM in the area of information management include Clarity (financial governance), PSS (e-discovery), OpenPages (governance, risk, compliance), Netezza (data warehouse appliances), SPSS (predictive analytics) and Cognos (business intelligence). How is integration of these portfolios coming along?
MM: We have a very robust platform that we built almost seven years ago to enable us to easily onboard our acquisitions. For example, SPSS we acquired and a year ago we closed the deal. And today with Cognos 10 we showed deep integration with our business intelligence environment so that people can use the SPSS as well as very tight integration with the modeling environment. We allow people to use tools on their own if they want to. We don’t make people use the entire environment but we onboard the capabilities in a way that makes sense to the customer. OpenPages for example (prior to our acquisition) had already embedded Cognos in their solution, they OEMed Cognos and had for many years. They surfacing all this risk data in a Cognos dashboard so integration with OpenPages is already done for that part of it, and there are other things we’ll probably do.
CWC: What role does the cloud play in IBM’s analytics strategy?
MM: Some things can be deployed to the cloud because they are activities that are done in the same way. Whereas with analytics, what a lot of people want is the solution deployed in the cloud for their own use but they don’t want their data necessarily to be up in cloud because they’re worried about security. They also do things differently; they may use the same data source but have a completely different model that they’ve built against it. It’s a lot of operational activity. When you think of a lot of the applications that are really thriving on the cloud like a Salesforce, it’s a prescriptive application that’s they’re putting up there that helps people manage their sales interaction. But people do it in a very similar way. But business analytics tends to be done based on how people think at that company, their culture, their view of the competition, their view of what the future needs to be. It’s always different from company to company even in similar areas of the business. We have identified some areas that have commonality. Certain reporting against very specific types of files like general ledger, accounts receivable, accounts payable, some human resources reporting. Those can be standardized, … customized but it’s going to be similar from customer to customer.
CWC: What sorts of hosted analytics offerings can customers expect in the near future from IBM?
CWC: With the advent of greater compute power and cheaper storage, how will IBM differentiate its analytics offerings for that changing landscape?
MM: I think there are some platforms that are commoditized but I think in general there’s still a huge differentiation. Where IBM has been focusing is in the optimizing for specific workloads because that’s the big differentiation. You can have a box and that box can run anything. But if you can have your engineers focused on making that box tuned to be optimized to deliver analytic workloads, and then you have your software developers optimizing for that particular hardware you’re coming at it from both sides, you can dramatically improve the performance and the customer experience because they don’t have to then go do that. Most people get a vanilla-based box, then they put different workloads on it and must optimize that hardware for every workload that they work with. The idea, here, is out of the box you’re going to get the hardware optimized to run the software and the software is going to be optimized to be sitting on that hardware platform. It’s also a differentiation for the hardware platform because we can sell analytics to anybody on any platform; we are completely open. But if we can say to a customer we’ve done work to optimize it specifically to run on the IBM hardware it becomes a more compelling value proposition. IBM has always been open and heterogeneous, but integrated across the stack.