But BI will only succeed if it is properly integrated. By Nick Groves, senior manager and Eduard Gracia, director in consulting, Deloitte.
Business information has never been so important and valuable to the chief financial officer. This has been brought about in part by the dramatic surge in data requirements driven by legislation such as Sarbanes Oxley, Basel II and IFRS.
But legislation is not the only reason for businesses to get their data in order. The explosion of data volumes, demands for information and the proliferation of increasingly sophisticated vendor tools across organisations have all contributed. The recent focus on delivering strategy through integrated performance management techniques and bringing together financial and non-financial data across integrated planning, budgeting and reporting processes, has led many organisations to look at business intelligence (BI) and ask if it can help address some of these issues.
BI solutions aim to automate and rationalise these processes. Yet many implementations fail, or succeed only by incurring disproportionately high costs. Part of the problem is the way the solutions are executed. More often than not, the broader context of the challenges is not fully considered at the outset.
Not linked to value
One of the main issues is that the majority of today's BI efforts have no clear link to business value. As a result many BI systems become the preserve of a small number of specialised analysts within a department. To support decision-making, BI needs to be relevant to the major strategic issues in the business (in other words, linked to value drivers).
Technology- not business-led
Too many of today's BI efforts are undertaken as pure technology deployments. Often if there is not enough emphasis on the people and process aspects of BI, it cannot become embedded in the core business processes of an organisation. It then risks becoming an ancillary bolt-on used only by a few ‘techies’.
No information control
The creation and management of important business information is at best an ad-hoc activity. Vital information on customers, products, suppliers and business geographies often exists in multiple locations, seldom tallies between different systems and is generally allowed to grow and evolve with no proactive management.
Data quality
The success of many BI projects is also often compromised by the poor quality of the data. While this is not a failing in the tool itself, it can be seen as a failure of the business to address key data management activities. Key processes and controls must be put in place to ensure that data is captured consistently and accurately at source.
So how can these issues be addressed at the outset of a BI project to deliver valuable business intelligence?
Link BI projects to value generation
BI projects must be selected and prioritised based on their explicit linkage to delivering value in an organisation. Too many BI projects are undertaken because the data exists, the tools exist and it is believed that doing the project will shed new light on some dark corner of the business. Projects should be undertaken because the outcomes will contribute directly to the generation of business value.
Balance focus on people, process and technology
The successful deployment and long-term adoption of a BI solution is dependent on it becoming an integral component of the business system architecture. This cannot be achieved by a technical team alone. The people and change aspect needs to be addressed, including the definition of roles and responsibilities for BI capability, development of training materials and embedding the system within the organisation’s support processes.
Also, business processes supported by BI need to be addressed with a keen eye on the relevance of analysis. Just because a data store can scale to accommodate vast volumes of data it is unlikely that an obsession with detail will lead to improvement in cycle times.
Integrate with operational systems
For BI to continue maturing and evolving, it has to be embraced by a wider community of business users. Key to achieving this is for BI to become more real-time. This can only be achieved by formal integration with the major operational systems and reference data sources in a business.
Improve data quality
BI systems rarely generate new information; instead they manipulate data from other sources. Hence their success or failure depends heavily upon data quality. Adequate resources need to be devoted upfront to designing the overall system so that the data is fed from the right sources. This avoids manual data manipulation and parallel data flows (for example, for statutory and management data) that could lead to reconciliation issues.
Information is repeatedly heralded as a major asset. Organisations that move away from applying the spreadsheet mindset and towards BI projects will reap the rewards.