Category: Analytics
Quicken Loans Inc. is the nation’s largest online mortgage lender and second largest retail mortgage lender. The company closed $140 billion of mortgage volume across all 50 states in 2013–2014. Quicken Loans generates loan production from Web centers located in Detroit, Cleveland, and Scottsdale, Arizona. The company also operates a centralized loan processing facility in Detroit, as well as its San Diego–based One Reverse Mortgage unit. Quicken Loans ranked highest in “Customer Satisfaction for Primary Mortgage Origination” in the U.S., according to J.D. Power for 2010–2014, and highest in customer satisfaction among all mortgage servicers in 2014 and 2015.
Quicken Loans’ Lock2Close model helps minimize potential losses resulting from interest rate changes, by accurately predicting closing loan amounts. State-of-the-art data warehousing and monitoring allow the model to provide business-driving, accurate insight. This model supports investment decisions on billions of capital market dollars daily.
The company has transcended model-centric predictive analytics to develop a “model ecosystem”—a suite of technology in and around the predictive model supporting accuracy, stability, and visibility of model function.
Mortgage lenders experience interest rate uncertainty from the moment clients “lock” their rates. In essence, hedging against losses due to interest rate changes involves finding a future buyer willing to accept the locked interest rate after the loan closes. To manage these transactions, hedge desk traders use the Lock2Close model to determine how much to buy or sell on the forward market to minimize the loss.
In a volatile interest rate environment, inaccurate model predictions can be costly. A holistic approach to model function has significantly improved the client experience of the hedge desk across multiple dimensions. Hedge desk traders gain immediate understanding of how the model output was generated, and the ability to dig deeper is automated as well. The ETL processes feeding the model have been optimized, leaving more time for fixing problems. Finally, a separate forecasting model allows the analytics team to identify problems in model accuracy as early as possible. The net result is the protection of $100 million of mortgage loans each day from potential losses due to interest rate fluctuation.
Category: BI on a Limited Budget
Firmex is a Deloitte Fast 50 technology company based in Toronto, operating since 2006. Thousands of companies choose Firmex SaaS solutions to securely share their highly confidential documents for corporate and financial transactions, M&A, compliance, litigation, and procurement. Over 75,000 companies worldwide have used Firmex virtual data rooms (VDR), including financial and legal intermediaries, investment funds, and corporations.
The goal of the Firmex Business Intelligence data warehouse (FBI DW) was to provide an accurate, reliable, and detailed view of product usage data for customer-facing, marketing, and product teams. Firmex wanted to make strategic and operational decisions based on hard data over theory. In order to complete this goal, Firmex needed to design and build the company-wide FBI DW and create a self-service BI tool to provide access to the DW.
The company built and implemented this self-serve analytics area on a limited budget in six months, and it vastly impacted both product and company.
Impact:
- Account management: Average monthly upsell amount rose by 188%. Overages detected through the BI tool have increased total revenue by 3.8%.
- Product: Data now informs team decisions on building and fixes.
- Marketing: The company communicates reliably with 40,000 users.
- Customer service: Support improvements are in preparation.
Maturity: Exceeded executive team expectations; DW will continue to evolve and impact the organization as it grows.
Relevance
- Organizational strategies
- Cross-functional design and development team: Rather than building a large team from the start, the company hired one resource with expertise in core BI/DW areas. To stay slim, the team utilized its own talented employees to collaborate on the new FBI DW.
- Self-service BI
- BI department and super users
- Scope management strategies
- Focused on cross-departmental metrics, high value, low to moderate complexity
- Phased/agile approach
- Design strategies
Innovation and creativity were required to extract data from 5,000 databases, and in some cases 30,000 databases, with the same schema.
Category: BI on a Limited Budget
Wells Fargo & Company is a nationwide, diversified, community-based financial services company with $1.7 trillion in assets. Founded in 1852 and headquartered in San Francisco, Wells Fargo provides consumer and commercial financial products at over 8,700 locations, 12,500 ATMs, and on the Internet worldwide.
The solution provides a multi-faceted approach through 1) a view of problem tickets related to network issues possibly impacting any community bank; 2) a view of incentive compensation goal acquisition; and 3) a view of customer satisfaction scores.
Wells Fargo’s “business-first” solution laid the groundwork for a balanced view of business-specific performance, pulling together interconnected relationships between various data points. If a network fails at a community bank and customers cannot be served, satisfaction rates suffer and employees miss their incentive compensation goals. Typically, such a multi-faceted business approach requires large capital outlays, but Wells Fargo utilized existing software, used systems that had been configured, and connected to already-modeled data. The result: users only need a browser and appropriate credentials to access the information.
With an Oracle data mart and ETL solutions pulling from various source systems, Microsoft BI was employed to create a semantic layer and presentation layer components, giving users a guided analytics approach with dashboards while providing self-service opportunities. The development methodology mixed traditional BI delivery with various agile approaches resulting in user-story-driven business requirements. Using a consultative and collaborative approach with our business partners, the solution was developed in short cycles to produce value earlier than typically seen at Wells Fargo. The business could partake in some portions of the development using tools already available, and the foundation is in place for continued advancements in 2015 and beyond.
Category: Big Data Technologies
TTNET was founded in 2006 and has become one of Turkey’s leading Internet companies. The company plays a pioneering role in the communication industry with its mix of digital products and services and also in developing responsible business models. Offering a combination of three components for communication technologies—Internet, IPTV and voice, TTNET meets all of Turkey’s communication requirements to generate digital convergence for audiences within different fields including education, music, and digital game platforms. Türk Telekom owns a 100% share in broadband provider TTNET.
The business purpose of the project was to help electric utilities by identifying an electric outage and pinpointing its location by analyzing Internet radius data in real time. The system provides information to find the location of a fuse or breaker that opened upon failure in the restoration process.
Power outage and restoration management is a top priority for utilities. It is critical to respond safely and efficiently to restore service to large numbers of affected customers. Utilities have been leveraging technology to improve outage management. The purpose of the project was to help electric utilities by identifying electric outages and pinpoint locations through the innovative use of the data correlation framework of big data analytics using remote authentication dial-in user service (RADIUS) protocol data.
The main benefits of this system are reduced outage duration, faster restoration, improved customer satisfaction (thanks to faster outage restoration process), and meeting the key performance indicators of regulation: reduced OPEX costs and improved workforce of utilities.
Category: Big Data Technologies
- Verizon (co-winner)
- Solution Sponsor: Teradata Corporation
Verizon Communications is a global leader that provides broadband as well as wireless and wireline communications services to consumer, business, government, and wholesale customers. The company also provides converged communications, information, and entertainment services over America's most advanced fiber-optic network, and delivers integrated business solutions to customers worldwide.
Current big data and analytics technologies provide the opportunity for companies such as Verizon to improve its relationships with its customers. Using innovative approaches to big data, combined with the use of analytical tools, Verizon has enhanced its customer relationships and with VBAP—Verizon’s Big Answers Platform, which is a hybrid Hadoop-Teradata enterprise data warehouse architecture.
VBAP simplifies discovery of information while preserving customer-sensitive information and customer privacy. These tools streamline customer interaction while providing efficient strategic guidance to customer-facing agents. The tool also leverages what is known about the customer to improve the overall customer experience, by reducing the need for repeated call-ins, reducing call time, and right-sizing the customer’s account, while improving close rates and churn. This approach captures Verizon’s “simple, smart, connected” strategy.
VBAP integrates and makes available business data sourced from corporate applications that capture billions of events each day and provides instant insights into customers such as previous service calls and past call history, service opportunities, promotional pricing and end dates, and competitive intelligence.
As a result, customers avoid multiple touchpoints by using intelligence based on customer interaction and issues, with fewer handoffs among departments. Usage and interaction data is intelligently stitched together to create dynamic guidance for call center reps and reduce average call handling time. Data related to product usage allows for system-generated recommendations so customers can choose custom packages tailored to their needs. Finally, actionable information based on insights from big data has enabled Verizon to increase upsells, reduce downgrades, and better control disconnects. Reps now know everything important about a customer to enable a quality, personalized experience on every call.
Category: Emerging Technologies and Methods
The Guardian is an award-winning British newspaper founded in 1821 with editorial offices in London, New York, and Sydney. It has 127 million unique browsers a month.
Prior to 2013, The Guardian had many disparate data systems but no integrated view. It couldn’t link online behavior to content to revenue. Data silos also meant that only data owners had access to their specific data; there was no central enterprise data availability. The lack of focus on Guardian data threatened to put it at a competitive disadvantage. To value its journalism in the marketplace, it needs to understand its readers, content, and advertisers.
The Guardian Executive Committee realized the product and commercial value in its data. Once an integrated warehouse is built, it can be used to create a more dynamic website through data-driven tools such as personalization and recommendation. Also, by leveraging what The Guardian knows about user behavior and preferences, it will be able to feed its own ad-targeting solutions and optimize ad sales.
In June 2013 the company decided to build an enterprise data warehouse, named the Ophan Analytic Warehouse (OAW). The warehouse architecture is cloud-based to allow flexibility and reduced cost. The company leveraged cutting-edge technologies to maximize performance and impact including an open source, scalable, real-time ETL, and the use of both a cold non-structured data lake and a relational model in Amazon Redshift.
The Guardian now has a full picture of its content, audience, and revenue, supporting enterprise decisions, optimizing ad revenue, and feeding into product design and development. The new data warehouse integrated its data for better internal decisions and KPIs and now provides a platform for commercial advantage and data-driven site components. A data science team has been built and is providing advanced analytics to improve understanding of The Guardian’s audience and their consumption of journalism. Having an integrated data warehouse has also prepared The Guardian for advanced Web technologies including personalization, social optimization, and rapid data-driven development.
Category: Enterprise BI
VMware is a leading virtualization software company. Its technologies simplify IT complexity and streamline operations to help businesses become more agile, efficient, and profitable.
The project empowered self-service business intelligence (BI) within a managed enterprise environment at VMware. Enterprise BI has been central in empowering business units and stakeholders to make faster, smarter business decisions with trusted data, transforming the company both culturally and technically.
The journey began by recognizing that to keep up with the increased demand for information across the enterprise, VMware needed to ensure it could scale to support its rapid growth. It formed an organization dedicated to enterprise information management (EIM), and aligned BI to deliver value. It embarked on a quest to empower users to serve themselves and develop their own reports and analyses from a single source of truth using the right tool to meet business needs.
To address the central issue of trust and confidence in data, EIM created a master data environment and identified business owners accountable for the stewardship of the data, which ultimately flows into the BI environment.
A centralized EDW architecture and enhanced visualization toolset was deployed in March 2014. Arguments about data quality have been replaced with data-driven decision making. To continually support the dynamic information needs across the organization, their focus is to find the sweet spot, allowing the business to innovate using the same set of core information, and drive insight into action, within an agile environment.
The power of executive influence has been invaluable in driving adoption. The executive teams share mobile visualization of key metrics during business meetings. Key practices contributing to the success of the project include a laser focus on the business, balancing control and agility, and continued flexibility to evolve as the business grows.
Category: Enterprise Data Management Strategies
VPBank is a fast-growing joint stock commercial bank in Vietnam. Established in 1993, VPBank has almost US $8 billion in assets and aims to be among the top three banks by retail and top 5 by overall business by 2017.
To set up world-class data management and BI capabilities, VPBank’s data committee, chaired by the CEO, established a business intelligence competency center (BICC) in early 2014. The core unit of the BICC, the data governance (DG) department, had two immediate tasks: manage data needed for immediate reporting/analytics and implement a bank-wide data governance framework.
The data committee realized that to have strong BI, it needed governance to ensure that data—the backbone of any BI infrastructure—was managed as a corporate asset. The DG department engaged consultants for 10 weeks to review the bank’s initial version of its data governance framework against best practices, draft key data policies and standards, and finally create an implementation road map.
The bank’s implementation approach hinged on several best practices in data management: executive sponsorship, change management, benchmarking with industry models, and quick wins. At the end of the first year, the implementation was on schedule, with substantial benefits for VPBank.
Application of data quality policies along six dimensions (validity, uniqueness, completeness, consistency, timeliness, and accuracy) utilizing 86 business rules improved data quality over 200 percent. More important, data quality was fixed in source systems to avoid future issues. Thanks to a comprehensive business glossary the bank created, discussions about interpreting report numbers were cut by almost 90 percent. The data issue escalation and resolution processes were streamlined, bringing real value to the day-to-day lives of people working with data. Initial outcomes of the DG program are quite encouraging, and VPBank will continue to focus on the path culminating in a self-sustaining stage of governance per the three-year road map.
Category: Enterprise Data Warehousing
Since 1994, SquareTwo Financial has worked with more than 2 million individuals and small businesses in North America to provide practical solutions to outstanding financial obligations, allowing customers to improve their financial position and become active participants in the economy. By removing the burden of debt, the company gives hundreds of thousands of customers each quarter the chance to regain financial stability, secure a new job, or even take steps toward owning their first home.
To ensure data agility for the business stakeholders, SquareTwo Financial developed an enterprise data warehouse (EDW) utilizing industry best practices. As a result, legacy databases have been decommissioned and the company now has a single source for analytics and decision making. The EDW was architected based on the company’s business process model, enabling business and data alignment. In addition, agile development is leveraged, delivering results to the business more rapidly and frequently.
The EDW has allowed the company to deliver on several innovative initiatives, including the incorporation of analytics into their transaction processing system and the ability to update business-driven data quickly and easily. Overall, the company now has more data agility, with relevant information more accessible than ever before.
Category: Government and Non-Profit
Founded in 1861, the University of Washington (UW) is one of the nation’s premier public universities dedicated to research and teaching. UW includes three campuses, a world-class academic medical center, and five hospitals. It awards 15,000 undergraduate, graduate, and doctoral degrees annually.
To support better data-driven decision making and strategic planning, UW Information Technology partnered with the Office of Planning and Budgeting to launch UW Profiles, a set of 21 Web-based institutional dashboards that allow easy access, exploration, and understanding of core data. Powered by Tableau Software, these visualizations are hosted inside the UW Profiles portal, which has a searchable catalog, robust documentation, and training on each dashboard. UW Profiles portal provides access to valid, defined data from the enterprise data warehouse (EDW), building on EDW enhancements that include creating standardized data definitions and integration of data across UW subject areas.
The UW Profiles portal lowers the barrier to effective decision making and places UW at the forefront of higher education nationally in using data visualization to provide a core set of key metrics. Data can now be explored and aggregated at any level of the institution—university, campus, school, college, or department. According to former UW president Michael Young, “UW Profiles is a one-stop shop for basic analytics about our students, their academic progress, and our university.”
Category: Performance Management
USAA is a member-owned Fortune 500 company that serves more than 10 million members of the military community and their families with a wide range of insurance, banking, investment, and financial services.
USAA developed a member service representative productivity dashboard to translate corporate strategy and goals in key measures and metrics, as well as track performance and productivity from the organizational level down to the individual representative. The dashboard shows individual results against an average of a peer group, and it uses a variety of data visualization techniques to best show achievement of key objectives and goals.
The development of member service representative productivity analytics was the culmination of a vision that enabled USAA to venture into new conceptual and strategic designs for performance and productivity metrics. The team was able to utilize new development methodologies, technical approaches, training, and change management practices to set the example for more business intelligence development across the organization.
The co-location and collaboration of business and technical resources resulted in a high-functioning team able to quickly deliver quality results. The project team delivered new data, tools, and analysis to the flagship insurance organization, resulting in new insights that were either unavailable or difficult to achieve in the past. In so doing, the solution has directly supported USAA’s corporate mission to deliver the best products and service to its membership.
Category: Right-Time BI and Analytics
- Uninor (Telewings Communications Services Pvt Ltd)
- Solution Sponsor: Teradata Corporation
Uninor is the India subsidiary of Norway-based Telenor Group. It has more than 48 million subscribers in its six commercially available circles: Uttar Pradesh (West), Uttar Pradesh (East), Bihar (including Jharkhand), Andhra Pradesh-Telangana, Maharashtra-Goa, and Gujarat. The company offers prepaid GSM services, and its current network footprint covers more than 50 percent of India’s population. Uninor has positioned itself as the sabse sasta (most affordable) telecom brand. Its three-pillar strategy of being the best in basic services, mass-market distribution, and low-cost operations has helped it reach number four in customer market share with nearly 11% of subscribers in its six circles.
Uninor implemented an enterprise data warehouse (EDW) and single business intelligence ecosystem to provide robust, scalable, end-to-end data integration from various sources into a unified state-of-the-art architecture. This delivered a single source of true data by providing 360-degree analytics on all customer touchpoints in real time. This made the entire system intelligent enough to enabled fact-based decision making. All this facilitates faster time-to-action in a highly competitive market.
The implementation has helped Uninor with a real-time self-care app available through all channels. It provides customer-related information such as usage, VAS, and data at a single place and has cut call center interactions by 80,000 every day.
It also offers a customized FMS solution for continuous monitoring and detection of unusual behavior patterns. It correlates data from multiple network sources and layers to avoid revenue leakage. With this solution, Uninor was able to offer an industry-first initiative, which compensates customers for dropped calls on the Uninor network (whether at home or while roaming).
Category: Organizational Structures
Award winners were chosen by a panel of independent judges who have expertise in BI and DW.