Banking on business intelligence

Those who want to survive and thrive have to move fast to keep up with customer demands, changing regulations and the risks of everyday business.

Business units and IT departments in the financial services sector are striving to meet these demands and balance them with increasing costs, tighter budgets and controlled risk, but success is possible for companies that can make sense of the vast quantities of data at their disposal.

As an increasing number of financial institutions are finding out, business intelligence and data warehousing solutions are providing valuable insight into a host of hidden information treasures. But how can organizations ensure they’re getting bang for their BI buck? FST spoke to three industry insiders to find out.

Stuart Frost is a well-respected high-tech entrepreneur with more than 20 years’ experience in the technology market. Prior to founding DATAllegro in June of 2003, Stuart led and assisted a number of companies to develop business strategies and raise funding. Stuart is best known for his role as founder and CEO of SELECT Software Tools. Founding SELECT in 1988, he led the company through rapid growth and expansion, including a very successful IPO in 1996. Over the course of the last 12 years, he has led multiple successful rounds of funding involving VCs, corporate investors, private placements and public offerings in both the US and the UK.

Richard Daley, CEO of Pentaho Corporation, has held key executive management positions in the business intelligence software market for over 15 years. Prior to founding Pentaho he was the Vice President of Enterprise Performance Management (EPM) at Lawson Software, as the result of the acquisition of Keyola, where Richard was CEO. He previously served as the Vice President and General Manager of the Analysis Tools Business Unit for Hyperion Solutions as the result of the acquisition of Appsource Corporation, where he was also CEO. Richard started his career with IBM where he obtained a solid foundation in sales, consulting and technology.

As Chief Executive Officer for Panorama Software, Eynav Azarya is responsible for the strategic vision and direction of the organization. Prior to being appointed CEO, he was VP of International Sales and Partnerships, overseeing revenue generation and international partner relationships while driving customer success. Before joining Panorama in 2001, Azarya was Chief Operating Officer for BrowseUp Inc., creator of a collaborative web publishing tool that supports knowledge management processes, where he led worldwide company operations including R&D, finance, sales and marketing. Azarya has also served as campaign manager for the Israeli Labor Party and as the CEO of the International Center for Peace in the Middle East.

How can business intelligence (BI) technologies help financial institutions manage risk, detect fraud, leverage customer insights and gain visibility into their profitability?
SF. In general, BI technologies are already widely employed in these areas. However, the quality of the resultant reports is often limited by a lack of access to detailed, high volume data. The new trend towards high-performance data warehouse appliances that can handle huge quantities of raw transaction data is starting to alleviate these problems.

RD. For those who understand how to leverage BI – welcome to a whole new world, where the traditional constraints of costs and architectural limitations in BI have been eliminated. Thanks to open source, BI is becoming standards-based and far more cost-effective, and now the only barrier to successful application of these technologies is the organization’s ability to execute on BI projects. Now a large portion of the dollars they used to spend on software licenses can be used to bring in external consultants with deep, vertical BI expertise to teach companies where and why to apply BI technologies.

EA. Companies that are successful in maximizing opportunities and mitigating risk have changed their management approach. They have adopted business intelligence and performance management solutions and tools that enable them to be proactive – in identifying issues, in making decisions, in executing on them and in evaluating their impact.

Even though there has been success with BI in the financial services industry, it has been primarily at the departmental or line-of-business level. What potential is there for enterprise-wide BI deployments within the financial services industry? And what benefits could this bring?

EA. BI has traditionally been a departmental initiative. However like in many other industries, financial services companies are faced with the challenge of keeping up with the growing number of decisions that have to be made on a daily basis. The complexity of corporate decision-making stems from the paradox between data volume and the human ability to act on it. Data keeps growing at an alarming rate, but the capabilities of the average worker to process the information inherent in the data haven’t changed.

Today’s BI solutions enable decision-makers to target specific issues they know about or ones they discover along the way. However the key here is to garner that knowledge without spending precious time sifting through mounds of data. We believe that this is best accomplished if the critical information a decision-maker needs is proactively delivered to him, allowing that person to spend time translating it into knowledge. By using that knowledge, people will know what worked and what didn’t, so they’ll be better prepared to decide and act in the future.

RD. Organizations need BI on an enterprise-wide scale so that every employee has the information they need about operations, customers, financial targets, and more to be fully effective in their job. But even as many organizations have had success at the departmental level with BI, recent research shows that cost remains the number one barrier for many organizations considering enterprise-wide BI deployments. The old BI vendors created departmental technologies with departmental pricing models, and they just don’t fit any more – not for enterprise-wide deployment. Have you ever talked to one of the big BI vendors about an extranet deployment? And they explain their ‘per user’ pricing model, and the extranet upcharge and the add-on administration licenses? Clearly, there’s a big gap between customer needs and traditional BI vendor solutions. Pentaho is here to fill that gap, and finally make superior BI technology available at a fraction of the cost compared to legacy offerings.

SF. The ability to analyze huge quantities of enterprise-wide data is enabling some financial institutions to generate very strong ROIs from risk management and fraud detection applications.

The regulatory environment for all companies has been dramatically affected by recent federal legislation such as Sarbanes-Oxley. How can better analytics aid financial institutions in addressing regulatory compliance?

SF. We view BI as a critical enabler rather than a simple aid in addressing regulatory compliance. In many cases, financial institutions cannot achieve full compliance without access to the raw transaction data discussed above.

EA. Compliance with acts such as Sarbanes-Oxley requires fast and easy accessibility to critical financial and non-financial information. Panorama provides solutions that address the need for information immediacy. The proactive reporting and real-time information management Panorama provides enables stakeholders to have the information needed to improve their company’s accountability in key business processes.

RD. Recent changes in the regulatory environment have turned BI from a ‘nice to have’ into a ‘must have’. Consistency, transparency and timeliness of information have become critical for organizations trying to manage compliance. BI helps with all of these – providing dashboards delivering consistent metrics, supporting operational business processes with near real-time information, and providing visibility into critical information typically scattered across multiple, disparate operational systems. Organizations will continue to spend billions on compliance, and you’d better believe the big BI vendors want their share. Just because you have to comply with Sarbanes-Oxley doesn’t mean you have to comply with outrageous costs for BI.

What are the practical issues involved in integrating BI into the existing application infrastructures of financial institutions? How are these being addressed?
RD. Generally, BI technologies have done a good job at integrating with source systems, as well as portals, to the point that both are expected in any BI offering. What hasn’t gotten enough attention is business process integration, which is more complex, but also more valuable than data or portal integration. BI was historically an ‘after the fact’ application, used at the end of the month or the quarter to review historical data to understand operational efficiency, customer trends and sales productivity. But today, organizations are increasingly looking to make BI a part of the day-to-day, operational business processes, and this requires a different technology approach. Complex, inflexible metadata layers and persistent OLAP cubes create fundamental barriers to effective process integration. Pentaho provides the only BI platform that was built from the ground up around business processes, with an embedded process engine, to make it easy and cost-effective to deliver BI within an organization’s operational business processes. That means BI delivers value at the point of work for front-line employees, not down the road for a small group of analysts.

EA. Some of the practical issues of integrating BI into a company’s existing information infrastructure include data security and the ability to set permissions on data usage. Security of data is first and foremost with financial services institutions. Today’s solutions – such as those from Panorama – not only secure delivery of data, but can lock that data down to the cell level. We also provide sophisticated permission settings on data access to ensure that users can only view and analyze information that is relevant to their needs, even though this data could come from a much larger report or source used by many decision-makers at varying levels.

SF. Many large financial services companies have tried to implement new BI applications on top of an existing data warehouse infrastructure that is typically already over-loaded. Adding large volumes of raw transaction data to an existing data warehouse can bring the system to its knees and jeopardize the performance of existing applications.

Data warehouse appliance technology can help alleviate this problem by allowing companies to ‘divide and conquer’ the problem. In a typical scenario, an existing large-scale data warehouse can be front-ended by an appliance that stores tens of terabytes of detailed data. Fraud detection and risk management applications can be run directly against this data. In addition, rolled up summaries can be generated on the appliance and fed into the existing enterprise data warehouse. This approach has two benefits: it provides a high-performance ‘sandbox’ for the new applications, and also reduces the load on the existing environment.

What best practices should companies in the financial services sector employ when choosing and deploying a BI solution? How can this help with measuring the success of a BI implementation?
RD. It sounds simple, but it all comes back to the business requirements. Organizations should select and deploy no less and no more than they need to address their requirements. Instead of letting a vendor tell you that BI success fundamentally depends on the deployment of a dozen different modules and capabilities that happen to exactly match what’s on their price list, decide what your business users need and focus on that. Deliver successes early and often. This usually means in smaller increments, or at the very least, ‘starting small’ from a deployment perspective.

Beyond that, the nature of software evaluations has changed in light of the widespread acceptance of open source technology. For the first time, financial services organizations can go beyond RFP responses and demos and actually test and use software before they spend any money on licenses. They don’t have to wait until after the sale when the consultant shows up to understand what the software really does and doesn’t do. It’s so much less risky than the traditional demo, proof-of-concept, seven-figure-contract model.

SF. The BI and data warehouse markets are in the midst of a major period of innovation and disruption. In BI, the traditional standalone vendors are being attacked by database vendors such as Microsoft and Oracle. In data warhousing, incumbents such as Teradata and IBM are being disrupted by new, appliance-based technologies. As a result, yesterday’s best practices may no longer be the most appropriate. Financial services companies should approach this rapidly changing situation with an open mind to ensure that they take advantage of the new products that are now available – before their competitors beat them to it.

EA. The one issue we hear the most from companies we work with is around performance. As more companies look to provide BI applications to their employees, and even inside customer-facing applications, high performance and availability are crucial. BI calculations, especially those within the financial services industry can be complex. Doing analysis, generating different views of information and sharing information with others requires systems that can scale and provide rapid response times. We recommend that companies run performance benchmarks when evaluating a BI system to determine which system is best for the company’s needs and infrastructure.

What are your hopes and expectations for the next 18 months regarding BI uptake in the financial services industry? What trends/developments will drive market growth?
EA. Business intelligence is one of the top priorities for companies today, including those in the financial services industry. Greater competition and a company’s ability to stay ahead of its competitors come down to who can make and act on decisions the fastest.

SF. Based on current uptake by customers in the financial services sector, we expect data warehouse appliance technology to be a major trend in the next 18 months. This will enable a whole new set of applications to be successfully deployed while reducing overall infrastructure costs.

New legislation is also driving the need for organizations to secure their data. Until recently, securing data ‘at rest’, stored in a data warehouse, was not a viable option for companies due to the high cost and poor performance of available technologies. New encryption technology is now available that allows companies to encrypt data ‘at rest’ with minimal impact on data warehouse performance and at a reasonable cost.

RD. Without a doubt, the same dynamics of commercial open source that have significantly changed other markets will transform business intelligence. Financial services organizations have been among the most rapid adopters of the Linux operating system, the MySQL database and the JBoss application server. The same dynamics of superior technology at a small fraction of the cost of proprietary offerings will drive open source BI adoption in 2006, and the ‘big three’ will feel the pain as their customers demand better products and dramatically better value. Those vendors all support Linux and Apache, and some support other market-leading open source technologies like MySQL or the Eclipse IDE. The question for yesterday’s BI vendors is this: what business reasons drove your customers to embrace open source in other areas, and how can you think those same reasons don’t apply to BI?

Operational BI will also continue to increase in importance in 2006, as financial services organizations capitalize on the opportunity to put information in the hands of front-line employees, directly within the context of their day-to-day operational business processes.

Finally, while extranets are not new to BI, 2006 will mark an ‘extranet renaissance’, so to speak. Much as open source technologies like MySQL and PHP have rapidly gained popularity for extranet applications, open source BI will likely present a very attractive alternative for organizations looking to deploy or expand their extranet capabilities.