Interview: Dell’s Migration of Internal Analytics to Statistica

Print Friendly, PDF & Email

John K. Thompson DELLI recently caught up with John K. Thompson, General Manager of Advanced Analytics at Dell Software to discuss Dell’s migration of its entire internal analytics system to Statistica (its own platform). The company did this by developing a set of common key performance indicators (KPIs) on which to run its sales business, entering into co-development projects with the business and designing and deploying the business management system. The interview touches on how Dell used predictive analytics to streamline its sales, cut costs and transform its data landscape.

Daniel D. Gutierrez – Managing Editor, insideBIGDATA

insideBIGDATA: What was the primary driver for Dell to move its internal analytics system to Statistica, aside from the fact that Statistica is a Dell product?

John K. Thompson: Aside from the fact that Statistica is a Dell product and the significant licensing savings that come from using your own IP, as we transitioned to becoming a more data-driven business, we were seeking a more powerful, yet easier to use analytics software suite, one that could help us more effectively deliver analytics across the company. Statistica provides full support for the entire analytics lifecycle, including model management, real-time scoring and business rules. It also simplifies data mining, predictive analytics and the analysis of both structured and unstructured data. As our internal organization would be primarily responsible for driving the process, we wanted a tool that they could quickly learn, and we wanted to provide them with the resources they needed for a successful migration. To do this, we opened up data labs in our Enterprise Data Warehouse where the business could innovate in an IT-sponsored ecosystem. This way, our business partners were free to have a place where they could load their own data, explore new analytics and lead business operations.

insideBIGDATA: Can you mention any pain points along the way? Were there any difficulties making this transition?

John K. Thompson: Technically speaking, the actual transition to Statistica was very smooth—it happened in just under six months. One of our largest challenges was making the cultural change that’s required for any company – including Dell – to thrive in the modern data economy. Data and information has to be shared freely and governed carefully, and IT and lines of business must work together toward common business goals. In an environment where business and IT are required to operate on a united front, both top-down and bottom-up changes are required. One top-down initiative was pinpointing change agents throughout Dell’s executive and senior management teams who could help with the model and vision of the overall project, while helping enforce governance as needed. We leveraged IT developers to help the businesses develop the skills to create their own reports, while proactively supporting them in various needs and challenges. As a result of this effort, we helped break the culture of moving data to the users, and instead, by enabling self-service capabilities, we’re now helping bring users to the data.

insideBIGDATA: After final deployment of Statistica, what have you heard from the user base? Is there general satisfaction?

John K. Thompson: Overall, our user base has been extremely satisfied with the migration, and it’s helped us move toward a mature predictive analytics model. As an organization, we’re now spending much less time managing disparate datasets and much more time performing analytics to support better decision making and taking actions based on those decisions. To ensure those analytics processes were being adhered to, Dell recently launched a Decision Sciences Academy, which provides a consistent, standardized training model for using proven analytics methodologies and Dell’s tools and processes like Statistica. We also think the process we took to drive the internal migration really speaks to the success of this initiative. That process included developing a set of common key performance indicators (KPIs) on which to run Dell’s sales business, entering into co-development projects with the business, as well as designing and deploying the business management system.

insideBIGDATA: What were the main benefits to the organization? Were you able to assign any metrics to how well this move was received by Dell?

John K. Thompson: Our organization has been extremely pleased with the migration, and the results say it all. Nonstandard KPIs and metrics have been reduced by 50 percent, data is now being synchronized across more than 99 percent of records and users now have access to reports on their own without having to contact IT. On top of that, this entire effort has helped Dell boost IT and business productivity dramatically, while cutting its BI costs almost in half. Increased productivity and cost savings have enabled us to expand our analytics capabilities by more than 20 percent. As a result, this has allowed us to move faster toward providing self-service prescriptive and predictive analytics reports. On top of this, the initiative has increased automation of standard sales reports and dashboards by 40 percent, and users reported experiencing 60 percent faster response times for prescriptive and predictive analytics.

insideBIGDATA: Any plans for the future for Dell using Statistica for enterprise analytics?

John K. Thompson: In the future, we want to improve data quality and provide the business with a unified integrated Business Management System – the BMS enables our organization to better rationalize BI reporting and spending. Ideally, we’d also like to enhance Dell’s analytics capabilities. To catalyze this, we’ve created specific teams to accelerate the use of analytics to mine the data they are now housing in the BMS and data warehouses, to better identify patterns, trends, outliers and other information that will greatly improve decision making. Down the road, we’d also like to expand analytics to front-line sales and middle management.


Sign up for the free insideBIGDATA newsletter.

Speak Your Mind