CXO Asia Article: BI: know what you don’t know

Business intelligence (BI) involves the provision of management information and is designed to improve decision-making within a company. According to a Gartner survey in 2004, for the first time BI made the list of top 10 CIO priorities. So BI is increasingly becoming recognised as a business necessity, but why now?

Business conditions in a constant state of flux; sales patterns change from place to place and from time to time; suppliers change their delivery schedules and their prices; customers become more educated and therefore more demanding; balancing on this shifting terrain, business managers are expected to deliver steady earnings growth – these are all factors Dave Kloc, Country Manager for BI appliance provider Netezza, identifies as increasing the need for BI.

“Somehow, they must smooth out the bumps and anticipate the changes,” he continues. “Combine all this with the fact that organisations are collecting exponentially growing data that they need to analyse and understand in order to keep pace, or ahead, of super-heated competition.”

Dan Sapir, Vice President of Worldwide Marketing at enterprise data integration firm Attunity, tends to agree: “The need to access and analyse data is a most fundamental business requirement.” But the interesting point he raises is that: “Today’s BI is yesterday’s report generators and decision support systems”.

Is BI therefore just rebranding for age-old objectives? Chart Chai Chayavirabood, Associate Director for web services consultants Avanade Asia, believes there is a certain element of truth in this. “The need for BI has been around for a long time – as long as people have been pondering how to make the best use of limited resources in the face of unlimited demand,” he says.

This is a point picked up on by Dan Vesset, Research Director, Analytics and Data Warehousing, IDC. He points out reporting languages were available way back in the late 1960s. We then jump to the mid-to-late 1980s, which saw the entry of several new companies into the BI market with client server technology. But it wasn’t until the late 1990s that web-based architectures for these BI tools started to emerge.

Now in the 2000’s, an overwhelming advance in technological capability means the basic processes of BI are more efficient and less time consuming – software technologies such as web services, change data capture (CDC) and data federation (enterprise information integration, or EII) etc. And this is exactly the crux of what has taken BI to the next level.

• Specialised – balanced scorecards, corporate performance management systems
• Analytical s – e.g. executive information systems (EIS), online analytical applications (OLAP), data mining, etc
• Enterprise reporting

All of these rely heavily on data warehousing, a technique where data is extracted from diverse business systems, transformed into a common and consistent format and loaded into a single repository (i.e. the data warehouse), which is usually a database. Business end-users can then use business intelligence applications to access the data warehouse.

Its diversity puts BI hand in hand with systems such as ERP and CRM, which saw prolific implementation throughout the 1990s. “While operational systems such as ERP allow organisations to achieve operational efficiency by optimising processes, BI solutions provide analytical capabilities to end-users to highlight and investigate areas of concern, and (to a certain degree) suggest answers,” says Chart Chai. “In this way, the two compliment each other. Operational systems allow you to do things faster, the BI solutions help you to think faster. Combining the two allows users to do, think and consequently act faster.”

Peter Barnes, Director Solutions Marketing at ERP vendor SSA Global, agrees: “ERP was widely adopted in the 1980s and 1990s. Many organisations now have mature ERP installations that hold a large amount of data. BI involves unravelling this data, generating greater value from an existing ERP system and improving overall organisational performance, so better, more informed decisions can be made quicker.”

In other words, BI makes enterprise data actionable and uncovers trends and patterns that might otherwise go undetected. Managing a business on intuition, educated guesses or averages isn’t good enough anymore. To be successful, a company needs a foundation of accurate, current and complete information.

“The importance of thorough and accurate BI cuts across all vertical industries and applies to any organisation capturing volumes of data that needs to be analysed in order to make business-critical decisions,” says Kloc. “Retailers are using BI to analyse their loyalty programmes to manage inventory, cross-sell and up-sell. Telecommunications carriers use BI to identify and mitigate fraud, or determine which of their customers are at risk for churn. A delivery service might predict which vehicles are most likely to break down and where. A bank might use BI to identify customers who, based on their recent activity, are likely to transfer their account to another institution etc.”

And it seems everyone in an organisation stands to benefit from BI, including busy mid/senior management users who can effectively monitor the pulse of their organisation by monitoring key performance indicators. For instance, business analysts can use BI to analyse trends, spot irregularities and institute pre-emptive measures before it is too late, and operational users can use BI to access good, clean and accurate data to monitor operational matters.

“BI provides a helicopter view, allowing an organisation to view their performance from the top down, drilling into the unique data an organisation requires,” asserts Peter Barnes, Director of Solutions Marketing at ERP vendor SSA Global. “For example, an organisation could look at last year’s revenue and drill in to exactly what was sold, where in the world it was sold, when, and why. It also allows trend extrapolation, allowing any business to identify the longer-term trends within the marketplace. If a company appreciates and understands the market trend, they can extrapolate into the future. The information can then be provided in different formats, such as an interactive dashboard or OLAP (on-line analytic processing).”

With BI seeming to be an all round necessity, it just leaves one last point to raise – perhaps the US$64 million dollar question: what should a customer be looking for in BI and is there a perfect solution out there? And the answer, resoundingly, is no! Each customer is unique so, therefore, it would be foolish to take a one-size-fits all approach.

“There are no ‘perfect’ solutions from any one vendor. The end-user must be mindful of the different needs of their end-users ,” Vesset admits. “Executives might want dashboards, analysts may need OLAP or data mining, other information consumers may need static online reports or limited interactivity and so on. Organisations deploying business analytics should consider the decision-making processes supported by the software rather than implementing standalone tools.”

Instead, Kloc provides some practical advice as to what a customer should be looking for. “An organisation needs a data warehouse system specifically designed to handle complex BI analyses. The ideal system combines massively parallel hardware, software and storage, and is directly focused on providing optimal response times and scalability at the multi-terabyte level,” he says. “The ideal BI platform eliminates yesterday’s patchwork of hardware, software and storage, and enables optimised access to information. As business and technical demands continue to grow and change, this BI platform should be designed to scale with data size, scope and performance needs, delivering significant performance increases over existing patchworks. And it does all this at an affordable and predictable price.”

There is, of course, still room for improvement (or perhaps development), and the increasing realisation of the potential importance of BI brings with it an expectant future where the “possibilities are limitless.” In the short-term, however, Vesset believes BI will provide more data visualisation, more predictive analytics based on data mining and statistical analysis, as well as tighter integration of BI tools with transaction processing software and support for more scalability.

Barnes contends that the next steps are already taking place now through CPM (corporate performance management). “BI is concerned with the analysis of performance to support better decision making. CPM is about closing the gap between strategy and execution. It goes beyond BI to involve the planning and budgeting cycles. Corporate objectives can be broken down and into specific departmental or role specific targets and then disseminated throughout the organisation,” he adds. “For example, once the company has agreed its corporate objectives, it can then distribute these to the appropriate countries, departments, teams etc. This ensures each team is accountable for a specific target that supports the overall corporate objectives.

“CPM also allows an organisation to measure its actual performance against its planned performance. Where BI gave only half of the picture, CPM now provides the bigger picture, allowing an organisation to look forward to where they are going, and how they get there.”

Long-term, the convergence of BI and business systems will allow users to perform real-time analysis and react much, much faster to anticipate changes in consumer behaviour. “The internet is evolving and will continue to change the way we work, rest and play,” states Chart Chai. “This opens up opportunities for BI to be used in ways that we cannot even begin to fathom. We are in for an exciting ride!”

BI: know what you don’t know

Other articles by GDS Publishing