Being intelligent about business

Most heads of today`s corporates will agree that the business intelligence (BI) market has changed, and CTOs planning IT strategies need to take note.

In the past, the emphasis in this market was on front-end tools. Now that they have become commoditised and integrated with a variety of back-end systems, the focus on BI projects is shifting to a number of back-end issues. Industry experts say sorting out these issues should lead to:

* Better BI system performance
* Reducing analysis and reporting times
* Closer alignment of technology and business objectives
* Better quality results

IDC estimates that by 2008, in the Asia/Pacific area (excluding Japan), the BI market will be worth $382 million. Adoption will focus primarily on tools for end-user query, reporting and analysis software.

Gartner estimates that companies in 2012 will deal with 30 times more data than in 2002.

A successful deployment is usually long-term, ongoing, complex and resource-intensive.

Charl Barnard, GM, Knowledge Integration Dynamics

Performance management also demands that IT and business executives work more closely to define business processes to be automated, establish business rules for performance measurement and associate metric owners with corporate data for accountability.

However, this is proving to be a problem for companies that deployed data warehouses four or five years ago, because the back-end architecture that was appropriate then must now change. The best approach for them is to redesign the entire data warehouse and redeploy it. Not doing so can result in poor performance and poor quality information, says Julian Field, MD of Centerfield Software.

The short-term correction to a slow or unresponsive data warehouse has been to throw more horsepower at it. Even on the extraction, transformation and loading (ETL) side of BI, some local companies are taking up to 12 hours to load data into a BI front-end tool to begin analysing it.

Charl Barnard, GM of Knowledge Integration Dynamics, says a successful BI deployment is usually a long-term, complex and resource-intensive process that, once in motion, delivers substantial returns and business value-add throughout the life of the organisation.

BI deployment value to the organisation is immediate, even before the first data mart or query tool is installed. It is good for any organisation to consider data quality management and integration, the first steps in a BI project, in the project plan and the eventual processes.

Barnard points out that unsuccessful deployment can have negative consequences, including a waste of resources and time. This can have the following negative results for an organisation:

* No efficient BI capability
* Fewer resources than before
* Back to the same situation as before the BI deployment project
* A negative perception of BI management or service provider
* Eventually, the ending of the BI project

"Successful BI requires more than simply good ideas and the right tools," says Barnard. "It requires strategy that spans the entire BI project, which is driven by business objectives, from conceptualisation through to the end-user finally running queries on the data required. This culminates in cost-effective, efficient plans and processes required for information delivery."

Rodney Hood, Business Connexion`s business unit manager of business applications, says many BI implementations have not delivered value, because they do not measure and deliver information on the areas of the business that most directly impact on business performance.

In the past, the ICT department would implement BI, but it does not always have the necessary knowledge of the key drivers of the business. Today, companies are presented with a sea of data from a variety of sources. The objective of BI is to enable the business to extract value from its data, Hood adds.

To achieve this, he maintains, it is not enough to simply implement BI tools without taking the necessary steps to ensure they deliver the right information to the right people at the right time.

"The first step is to educate CEOs and senior management as to what BI can deliver. They then need to be engaged in the process so that their knowledge is harnessed to ensure BI is delivering information on key areas that affect business performance," Hood says.

Martin Rennhackkamp, strategic information director at Prescient Business Technologies, counts BI as a major corporate resource, on a par with money and labour.

"The top management structures in the large corporates in SA do not seem to have a clue of the importance of BI or information management. If they did, their organisations would not have three, four or in some cases even more than 10 non-converging BI initiatives being pursued concurrently," Rennhackkamp comments.

This is one of the biggest problems in the BI space. The BI information as well as the related processes, approaches, systems and required governances are poorly understood by the top layers of management and decision-makers.

Unfortunately that is exactly the prime audience that BI has to feed with crucial information from which they make day-to-day through to strategic decisions.

Yes, they understand the information and the value of the information delivered through BI; but do they understand the process of BI and the importance and value delivered through that process? In Rennhackkamp`s opinion, the answer is no.

Peter Mbelengwa of Siebel Systems SA echoes his sentiments, adding that the benefits of a clear understanding of BI can be enormous. In this era of closely competing prices, organisations running on almost stock-standard operational systems (like ERP systems), and in the tight legislative governance of some industries, the only edge you have over your competitors is in information - delivered to the strategic decision-makers through BI.

"It is like competing in a race with a stock standard motor car or windsurfer, where the difference between the winners and other racers is in their knowledge of the course and the conditions, and how to exploit it to win. The difference is not in the equipment or the products any more," Rennhackkamp notes.

Most companies interviewed for this report agree that architecturally, the best solution available with current technology is still the enterprise data warehouse, with optionally downstream information delivery marts, populated through an ETL tool and exploited using user-empowering query-and-reporting and analytics tools.

Here the old adage holds: keep it simple. The simpler and more streamlined the BI architecture is, the easier it is to be maintained, to be expanded and to adjust it to cater for changing business needs. The simpler it is, the more efficient it can be implemented as well, making things like near-real-time data warehousing practical realities.

Some corporates have the most elaborate BI architectures. You do not have to store the data in a normalised structure as well, if you are going to use appropriate dimensional structures for reporting and analysis. You don`t need to kludge an operational data store (ODS) into the architecture if you can get the data fast enough into the data warehouse.

Near-time decision-making information is being required more and more in the highly competitive landscape. The technologies are there to make it happen. It is mostly red tape, politics and other bureaucratic obstacles that prevent organisations from achieving these goals (that is, if they really need it). Instead of removing the hurdles, they then kludge an ODS into the architecture, creating another hurdle which then later becomes difficult to manage and even more difficult to remove.

Rennhackkamp explains that, methodologically speaking, the best way to deliver BI solutions to the business is through an iteration of short lifecycle (three to six months) sub-projects.

It is crucial that the business users and business sponsors perceive it as a series of independent, business-focused projects.

But back at the (implementation) ranch, these sub-projects must be tightly integrated around an integrated, conformed enterprise data warehouse data model, single extracts per source system, stringent data quality controls and data integration rules. It requires strong and skilled program management to get this balancing act right.

Ampie Swanepoel, a senior corporate performance management analyst at Dimension Data, says there is no question of the necessity of BI and corporate performance management (CPM).

However, many large corporations with big investments in BI, both in the form of software and data warehouse developments, have been disappointed by the lack of business value derived from these initiatives.

Organisations that have acquired sophisticated BI software tools are realising it requires a bigger effort than originally anticipated to deploy these technologies. The focus within these organisations is to ensure they derive value from their current investment in BI before they spend more on additional licences.

According to a recent Dimension Data white paper on BI and CPM trends, it is likely that more consolidation will take place in the BI technology space. It is anticipated that the large enterprise resource planning (ERP) vendors (Oracle, SAP) will continue to enhance their BI capabilities and might start to dominate the BI space within their customer base.

The best of breed BI vendors (such as Business Objects, Cognos, SAS, Hyperion) will still have a place, as it is not anticipated that the ERP vendors will not be able to cater for the needs of all organisations. Because BI is their core business they will continue to bring more innovation.

It is anticipated that the consolidation will come from within the smaller niche player sector.

It is clear that ERP vendors see a huge potential for growth for themselves in the BI and CPM domains. The market potential for ERP vendors selling integrated BI and CPM modules to their existing base is huge, which will help them gain significant market share. Furthermore, the ERP vendors have the domain knowledge, which spans across areas such as finance, HR, sales and marketing to deliver a comprehensive CPM suite. They also have the R&D facilities to develop a comprehensive CPM application in a short time.

Most organisations that have invested in ERP solutions have been disappointed with the BI and CPM capabilities. Although ERP systems have largely addressed the needs of operational users, they have not been able to address the needs of tactical and strategic users.

The main reasons they have not been able to do so are:

* The user interfaces of ERP systems are too complex for operational users.

* The closed architectures of ERP systems make it difficult to integrate data from external systems.

* Apart from organisations that are disillusioned by the ERP promise, there are many organisations that will probably never go the ERP route. For this reason, organisations will continue to turn to BI and other niche CPM vendors for a solution.

* Many BI vendors claim that CPM is the next big thing for BI. Although CPM suites and BI tools often have the same technology base in common, CPM also involves the processes, methodologies and metrics that are needed to manage the enterprise`s performance. In its current state, BI usually concentrates on providing analytic capabilities using graphs and reports. The measurement part is important for CPM - hence the overlap - but it is not all there is to CPM.

* Furthermore, CPM is not BI`s successor. CPM is a more strategic application of BI.

Most industry players say that in medium-sized businesses, BI is working well, because it is usually directly driven according to the organisation`s strategic objectives, and tightly managed and governed.

In the large corporates, there has been an interesting split. Banking, insurance and the telcos are extremely bad at managing their BI initiatives from top-management level and at putting the necessary governances in place. They cannot show a good return on their BI investments for this reason. In most of the large corporate retail organisations, BI initiatives are lean, mean, tightly managed and tightly governed. These organisations get good management information and a high return on investment (ROI) from the BI initiatives.

The BI initiatives in the other industries are too diverse to classify under such blanket statements - some do it right and get both the information and the ROI, others not.

Amir Lubashevsky, spokesperson for Magix, Saharan distributor for Panorama, says his company can testify to BI`s effectiveness.

"Our customers are the evidence. Nearly 800 companies worldwide depend on Panorama`s BI solutions to better leverage the valuable information stored throughout their organisation to understand business performance and make decisions that improve their results," he says.

However, Caron Mooney, director of IS Partners, a Microsoft gold-certified partner for BI, says many organisations are inclined to think that implementing a BI solution is a matter of choosing the right tools and slapping them on top of your database. This mind-set, she believes, will lead to failure.

Essentially, BI projects by their nature should be scalable and extendible, but this characteristic should not be exploited.

BI solutions and BI programs, on the other hand, should be extremely scalable and extendible, and this characteristic should be exploited.

A BI project should last three or four months, and be a focused implementation effort to satisfy a single business or departmental objective. By scaling the project, we would violate this. Instead of scaling the project, organisations should scale the program and the solution.

Scale the solution within its governances (single source extract, single conformed enterprise model, etc), but don`t compromise on the delivery of a particular business-focused project. Rather complete that project quickly and start on the next one, or start another project in parallel, than extend the project and never complete it.