Veera Narayanaswamy
VP and global practice head for data warehousing, BI and performance  management, Cognizant Technology Solutions

In the competitive and fast-paced world that is the IT services market, Veera Narayanaswamy has always insisted that he'd rather see his company be the best, rather than the biggest in the field. But as VP and global practice head for data warehousing, BI and performance management at Cognizant Technology Solutions, Narayanaswamy has seen tremendous growth and a shift from traditional IT services to much broader and more strategic engagements with clients. From a five-member seed team in 1997 to about 3,600 practitioners today, it's still all about creating value to become top of mind with customers, as he recently explained to DM Review Editorial Director Jim Ericson.

DMR: As BI strategies have evolved, how has Cognizant adjusted to changes in the demand for services?
Veera Narayanaswamy: We started out 10 years ago selling data warehousing as a concept rather than a technology, but we needed to access technologies like data integration and databases and put them into a value chain solution. It was more like consulting - we came out with a lot of "foot in the door" ideas in small capsules that took eight weeks or so. If a customer was looking to buy a BI or ETL tool, we'd take the requirements and evaluate all the tools and suggest which was best. We were able to get small engagements well within the approval budgets of the project manager or director. Today we no longer sell concepts. I feel a strong movement from information monarchy to information democracy where every information worker is looking at BI as his or her right, with the attitude that information should be available at all levels for day-to-day decision-making. This has the full attention of the CIO and puts tremendous emphasis on organizational maturity and readiness to bring about the kind of infrastructure needed. Customers are looking at reorganizing their BI service providers from a suite of BI services to a BI superstore concept. That means a single partner to provide them with end-to-end BI services and ensure thought leadership, consistency, repeatability and year-over-year lower total cost of ownership.

DMR: How did you ramp to the broader superstore concept?
VN: From a services standpoint we adopted a three-legged model. One was based on the domain standpoint where we created a lot of business-focused solutions. In supply chain, for example, we have a solution that creates retail analytics and also provides corporate performance management for manufacturing. A second leg or type of solution is from a profit standpoint, an enterprise BI service stack powered by industry tools, which we call Global Reporting Shop. It's a unique approach to help customers move onto an enterprise BI reporting platform and helps build a strong return on their planned or existing investment. And from a thought leadership and strategic advisory services standpoint, we have incorporated evaluation frameworks, BI investment strategy frameworks and the corporate performance management framework. Another framework addresses maturity, where we benchmark an environment against industry peers. And then we tell them their competency level is here, and their maturity level is here; if you want to get there, these are all the steps you need to do. Those three aspects make customers confident to work with us as strategic advisers and not just implementers.

DMR: How rapidly are customers adopting the BI competency center or center of excellence (CoE) model?
VN: The trend has been the growth in importance of BI within the larger organizational landscape of customers. We've moved from one-off requirements for BI or ETL to a new phase of demand for BI within customer organizations and business intelligence centers. One driver for this is the fact that BI is getting applied to many more enterprise-level issues, such as regulatory compliance, which require much more of an enterprise view. That's possible only if they start building out processes, systems, infrastructures and organizations that are more centralized. We've seen customer organizations building a central data warehousing organization and then creating what we call a business intelligence factory. A BI factory is basically a sourcing model or a people model where various downstream business departments can find their people in a centralized manner. With the CoE model, customers are coming back and asking for a different set of services on a large scale to build a development center around business intelligence.

DMR: We hear about this as a goal, but how many have achieved that state of maturity?
VN: Certain companies we work with in financial services, telecom industry and pharmaceuticals have actually been able to move toward that kind of model. Some of that is naturally fueled by other technology and infrastructure-level changes that need to happen. When people say they are going to move off 15 or 20 BI platforms and zero in on three or four technology standards, the downstream business departments immediately get short staffed in terms of technology skills because they have been on a particular set of technologies all the while. With that kind of movement, business has to come to IT and figure out a centralized source of BI staff. That tends to have a cascading impact on an organization's capability to put in place the CoE and operate it effectively.

DMR: That seems inevitable, but it makes me wary that we are recreating the old monolithic, slow-moving IT service model.
VN: You raise a concern that is valid for many downstream business organizations. When we introduce a factory model, we always need to help IT address the pushback. It never works to go in and say that a BI and DW CoE is going to be owned by corporate IT, who will manage the team, decide on the scope and decide how many people are required. That's the monolithic program you are talking about, but that is not our approach. Instead, we have a development resource pool of people for you to take and use in your own environment. The factory model allows business to choose from the pool as opposed to sending their requirements into some closed program. The business is more comfortable because it owns the pieces and knows it can make things happen and move its consultants to the speed it wants to. IT is more comfortable because it has the people governance and can maintain the pool with a good level of quality control. And of course we are happy because all those people come from us.

DMR: But companies still have to confront the supply constraints of BI staff.
VN: The business is no longer paying for IT projects; it'll only pay for business projects that have the ROI in seven to eight months or so. Rather than the big-bang approach, we'll start something small, build on it, use it as a springboard and see the success as a move to other areas. Even in the factory model, we are accustomed to having 20-25 people being utilized by two or three departments for a couple of months. After six or seven months, we ramp it up to 70 or 80 people, and at the end of the year, we'll typically have 100 to 150 people working on the CoE model. As for standardization, the middle path is also preferred by customers. Organizations are trying to consolidate at the business semantics level where they build a common metadata layer and define what we call an enterprise report library, basically a set of reporting templates, and let business departments down the line feed off this level of standardization regardless of the technology they have.

DMR: I have read about your concept of "data health." Could you briefly explain that to our readers?
VN: Well, I have been promoting something I call enterprise data clarity or EDC, which looks at data from an overall health perspective. We talk a lot about data quality, but more important is the strategic use of data, which is not determined just by quality. You need to look at the health aspects, the timeliness, the availability of data and so on. It's very important for people using data to have clarity of where it comes from, if it's the right data, how it is massaged or transformed from its source. I have come up with a fit-for-function scorecard as part of the EDC framework and have been giving speeches on the subject, and we are building EDC systems for our clients. I think it will be a major emerging trend.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access