Business Analytics and Predictive ITBy Susan Nunziata | Posted 05-05-2011
Business Analytics and Predictive IT
As vice president and CIO of IBM, Pat Toole focuses on advancing the company's transformation agenda and aligning IBM's information technology investments to the business strategy. Business analytics and predictive analytics play a big role in how his IT organization works with business stakeholders in the enterprise.
Likewise, making sense of massive amounts of data is instrumental in the day-to-day functions of Toole's global IT operation. While company policy prevents Toole from revealing his total IT headcount, he stresses that his organization's role is to understand the skills of the 400,000 IBM employees around the world, and how best to apply those skills to projects that will deliver real business benefits.About two years ago, IBM created the business transformation and technology shared services operation that Toole now oversees.
"We're organized fundamentally into three main missions," says Toole. "The first is to run the operations of IT. The second is to lead and manage transformation projects to drive better outcomes. And the third [involves] an integration team to understand the needs of the business." This is set up on a global, centralized shared-service organization. Toole leads a team of nine vice presidents worldwide and reports to Linda Sanford, senior vice president enterprise transformation.
In November 2009, the company announced Blue Insight cloud, the world's largest private cloud computing environment for business analytics. Blue Insight cloud consolidates information from 100 different warehouses and data stores, providing analytics on more than a petabyte of data. It's used to provide insight that helps IBM sales teams and developers meet the needs of clients worldwide. The solution is expected to deliver tens of millions of dollars in savings over a five-year period.
Toole spoke with CIO Insight Editor in Chief Susan Nunziata about the dynamics of his IT operation, how Blue Insight is influencing both internal and external data analysis, and the ways in which he sees business analytics transforming the role of the CIO in the next 12 to 24 months.CIO Insight: How has your two-year-old IT structure influenced the day-to-day running of your IT operation?
Pat Toole: We went from having service-level agreements (SLAs) for specific applications to having SLAs for business processes. That incorporated 15 process owners who cut horizontally across the organization. The important thing to us is to make sure that the processes are working and effective, and deliver the outcomes that are required.
The structure allows us to set up portfolio managers based on business process, so we can understand intimately, from the business person's perspective, what the meaning of success is. For example, if we were babysitting applications, and an application upgrade was undertaken that required the app to be out of service during a period when the business-side's needs were critical, then you would not be meeting the needs of business process.
We try to look at what we need to benefit business owners, what they need to have running, and when it needs to be running. Then we measure ourselves against those criteria.
How do you lead and manage transformation projects while still handling the day-to-day demands of IT?
Toole: It's essential for all CIOs to manage but also have initiatives under way to accelerate the business. We develop a list of transformation projects. Those tend to split fifty-fifty between projects to drive revenue growth and productivity projects to drive business processes and make the operation more effective and efficient.
That ratio has changed over the years. In general, shared services has focused on cost reduction. But we made a fundamental shift to look at what is required not only from [the standpoint of] productivity savings, but from revenue growth as well.
A simple example is being able to create software-as-a-service (SaaS) cloud environments, so clients who want to buy that way now have that opportunity. We've ramped up environments to support that. It's an efficient allocation of resources, and how you apply those resources. We've worked to understand what the market opportunity is, where the market is growing, what we need to do to understand the skills of our 400,000 IBM employees around the world, and how to apply those skills.
Who has a seat at the table when it comes to creating transformation projects?
Toole: Depending on the culture of your corporation, you have to have a strong governance model. We have a number of formal settings where we talk to stakeholders about priorities and business process and what's going to drive value.
We meet formally with process managers; we meet with business owners. We also meet with our peer shared-services organizations on a monthly basis. Each company has to take a hard look at the governance model and make sure there's a closed-loop system in place to prioritize processes and make sure the outcome of those projects is made manifest.
The CIO sits in a unique position to see the entire company end-to-end from a process perspective. So if you think about who is in a good position to build collaboration and strong governance, the CIO is in a position to do that. The CIO has access to the data and understands where all the data resides, and has fiduciary responsibility to shareholders to make sure that data is a trusted source.
We'll return to that question of managing data, but first tell us more about how your integration team operates.
Toole: It's another good example of a key component of our overall governance model. When we created shared services, one of the key concerns on the business side was whether or not IT would become disconnected from day-to-day business needs. Through our integration team, we have people who can participate on a day-to-day basis with the needs of business units and maintain good insight into what those business needs are. They can translate and articulate to the CIO's office what is needed.
We work very closely with the office of the controller to have a rigorous business case development model. What's more, we have an independent source to work with my team and a business unit to make sure that any benefits are clearly articulated in the business plan for that particular unit. It helps us across the corporation to have a uniform way to measure these projects.
Predictive IT: Building a Good Business Case Development Model
What advice do you have for other CIOs looking to build a good business-case development model?
Toole: Focus on hard benefits--things you can measure, versus things that are based on soft metrics. One thing we try to do diligently is harden up the soft savings. What does being more productive really mean? Does it mean that if we apply more resources we can generate more revenue because we're more productive? Or does it mean that we can shift people to other parts of the business to generate more productivity there to help increase sales? If you harden up benefits, at the end of the day the controller can really measure the outcome in a tangible way that comes back to shareholder value.
To quote Mark Twain, "There are three kinds of lies: lies, damned lies and statistics." So, how do you extract the right data to achieve your goals?
Toole: Really focus on data that drives value to the business, whether that's client-oriented, product-oriented or employee-oriented. Start there, and make sure you have trusted sources of data for your clients, your products and services, and your employees
Then the question is, "How do I get insight out of that data?" In our case, we noticed that people were using a number of different tools and warehouses to manage data. In response, we created Blue Insight, which combined 100 data warehouses totaling about a petabyte of data. We worked with Cognos and created a single tool that enables us to access that data. We now have 50 applications accessing data, and we are adding another 50.
It's about creating trusted sources of data so you can really do the processes you need to do. On the other side, you have to make sure that the data is as automated as possible and can be analyzed in actionable ways.
In our Blue Insight cloud, we now have 185,000 people who have access to that data, which we expect to grow to 200,000 shortly. It has 41 ported applications, and another 95 applications are in the pipeline for deployment toward our full-year user deployment goal.
What advice do you have for other CIOs looking to execute a similar undertaking?
Toole: When we created shared services, one thing we did was create centers of excellence on a few strategic topics; data was one of them. We created a center of excellence around business analytics that was tasked with inventorying what was in the environment and coming up, [then determining] ways to systematically consolidate it. That really helped us get our arms around a very large challenge.
Here's an example: small-deal management. This is when we're trying to take a business lead, analyze the lead and pass it on to the appropriate business partner. We were able to take a significant amount of time out of that whole process by applying analytics. It doesn't take a big investment to get a big payback.
Another example: We created a CIO dashboard, and now we're able to, in a single place, go and look at our entire operation--from what's happening in infrastructure, application portfolio management, and risk and compliance, to transformation initiatives, etc.--and analyze all the risks and benefits being delivered. It's allowed us to be much more effective in our own role of being a shared-services corporation.
The next thing we're applying is predictive analytics on top of the IBM Cognos environment. We do a very good job of managing infrastructure and reacting to any issues that come up. Now we're using predictive analytics to analyze our server log to see if we can identify trends and fix problems before they even happen. We're applying that same kind of predictive analytics to business ideas.
That's the next step: Taking data we have and predicting what's going to happen in the future.
We hear increasingly from CIOs about the business-side demand to have mobile access to analytics information from any device, anywhere. How is that affecting your use of analytics tools?
Toole: In today's world, with mobility growing as fast as it is, you have to think about three fundamental things: establishing a collaborative environment where people can work together regardless of location; within that environment you need to create communities for people to work together in intimate ways; and they must have analytics tools and predictive-analytics tools to enable them to work together as effectively as they can in these global collaborations.
We've put analytics around communities to understand how those communities work and generate best practices to bring all communities in our global environment to their optimal performance.
In building digital reputations, you have to be able to secure the environment by employing technologies, such as data loss prevention (DLP), and you have to do a lot of education so people understand how important data protection is.
There is a big role the CIO has to play, in a much broader sense, in [handling] all the potential security challenges that come with a global mobile workforce. That whole landscape of risk management is becoming a larger part of the CIO's role every day.
Where do you see the role of the CIO evolving in the next 12 to 24 months?
Toole: It's a great role to have, because you do get to see the company end-to-end. There's a real opportunity to deliver value.
Going forward, CIOs have a strong role to play in working closely with business leaders to generate revenue growth and productivity. There's so much market opportunity. There are so many exciting solutions that are changing how technology gets delivered--such as cloud technology. And there are so many benefits to helping the business get insights into data and then taking it to the next step to help them predict what will happen in the future.
The role of CIO is a critical one in the next 12 to 24 months, especially now that businesses are reprioritizing on both growth and productivity. We need to be leading the business, not following the business, in order to be truly effective in our roles as CIOs.