Skip to Content

Introduction

The termsbusiness analytics and business intelligence are often used interchangeably. Though they may at first seem similar, each has unique objectives and approaches. Business intelligence (BI) uses transactional and historical data to determine trends, analyze what has happened in a business, and report on outcomes. Business analytics, in contrast, creates a performance management framework and then uses data to drive predictive assessment, often via data mining, statistical modelling, and scenario analysis. And while the need for business intelligence is usually recognized and addressed, few organizations have yet recognized the need to integrate analytics into the corporate culture.

The result, if analytics are considered at all, is too often a fragmented and disjointed processes like that illustrated in Figure 1. Here analytical applications and business processes fail to interact – leading to both data-management and cultural challenges throughout the BI environment.

 

 Figure 1

Figure 1: Typical BI and advanced analytics architecture

Disadvantages of current analytic frameworks

Siloed applications and processes

Analytical applications are designed to answer specific questions, but often have scarce or tenuous links to other business areas – so output from an analytic application may not be reflected in financial forecasts or operations. For example, a campaign management application generates optimized campaigns for a particular product. The resulting increased demand, however, may not be linked to inventory management and forecasting applications – resulting in an out-of-stock situation on a particular promotional product.

And analytic insights may not link to a business’s strategic plan. For example, suppose an analytic application identifies a market opportunity for a new product, but that opportunity may not be incorporated into the strategic plan. Conversely, analysis might identify a promising new market opportunity for an existing product – at the same time the strategic plan is phasing out that very product.

Meanwhile, a lack of write-back mechanisms may mean that insights from analytic applications are not automatically captured in transactional systems, making it difficult to close the loop from insight to action.

Finally, with multiple, disjointed systems comes significant additional costs for maintenance, plus significant time and effort required to reconcile data.

Inconsistent data

A lack of consistent data definitions and business rules across BI and analytic applications – say from supply chain analytics to financial analytics – makes it difficult to compare and act on results. For example, a company’s sales and commercial departments may define “sales revenue” differently, compromising data aggregation and analysis.

In addition, analytical models and their results may not be consistently persisted in corporate data repositories, making it difficult to provide a historical record for auditing and corporate learning. When analytics are performed locally within a specific tool, models and results are not retained in the data warehouse for future reference.

Sceptical business users

Operating independently, siloed analytic teams can push insights to end users with little to no involvement from actual decision makers in the business. Without collaboration, such teams miss valuable input that could drive insight and generate buy-in. Meanwhile, business users are often sceptical of analytic results that they have not helped generate.

Figure 2 illustrates the impact of disjointed process on the business. BI provides answers to “what happened?” while advanced analytics suggests improvements and answers “what if” questions. Ideally, the two processes complement each other – but if disconnected, they could actually work against each other. To ensure smooth operational execution, a business needs an advanced analytic system built on, or in conjunction with, its BI framework.

Figure 1

 

Figure 2: Current business processes for using BI and advanced analytics

Benefits of an integrated framework

As a best practice, SAP recommends an integrated framework that combines analytical applications with traditional BI solutions, underpinned by a holistic performance analytics framework. As the example in Figure 3 shows, such an integrated architecture has several benefits – better strategic alignment, consistent data definitions and use, and business users who embrace analytics to accelerate performance.

Figure 3

Figure 3: Future state best-of-breed architecture for integrated BI and advanced analytics

Strategically aligned application architecture and business processes  

If you start with a strategic top-down blueprint, all your BI and analytics applications can be architected according to a coherent set of solution requirements, scope, and vision, factoring in all inter-dependencies. Such a blueprint could potentially enable drill down from a strategic KPI at the top level of a balanced scorecard to the lowest degree of operational detail – describing, for example, how KPIs relate to operational metrics at the user level. A top-down analytics environment permits close alignment of a company’s analytics architecture with its enterprise performance management and strategic planning processes. Adding advanced analytics to the mix could also potentially enable you to forecast your performance under a variety of business scenarios. 

The top-down approach also identifies and builds in interdependencies among business areas, helping to balance worthwhile, but often conflicting, objectives of different business functions. For example, a bank’s sales team may want to increase the number of new loan customers while its credit department works to increase loan portfolio quality – objectives that could easily conflict. An integrated suite of analytic applications can eliminate such conflicts by optimizing decision making across conflicting constraints, ensuring smooth operational execution.  

Meanwhile, an integrated feedback loop can ensure that insights derived from an analytic system are picked up by the transactional system, so those insights can be acted on. For example, customers identified as profitable with a promising future lifetime value can be tagged in customer service applications so they receive higher service levels, increasing satisfaction and leading to greater purchase and retention rates.

Consistent and more accurate data

In this model, all insights are acted upon – and all actions are analyzed. Unlike current analytical applications in which outputs and insights are not persisted, in an integrated approach, new analytic insights are written back to the data warehouse each time. This provides a historical base for past decisions, improving both traceability and auditability. New records also become assets to corporate memory, enabling future learning.

In an integrated environment, too, all BI and analytics applications use data from the same corporate data warehouse and operational data stores, ensuring consistent data definitions throughout the company. Data sourced from data warehouses also goes through standard cleansing and quality processes – improving data quality while drastically reducing both data inconsistency and lack of confidence in analytic outputs.

Collaborative culture 

The integrated approach enables a single umbrella team – such as an analytics competency center – to own all analytics and BI applications and initiatives within the organization, with sub-teams dedicated to specific business areas. This approach facilitates a culture that more closely integrates analytics into standard operational processes, while fostering collaboration between line of business users and the analytics team – to create analytics that truly improve performance.

In a best practice environment, business users interact with the analytics team to define the questions that need answers and validate analytic insights. Business users then feed the real-world results back to the analytics team, further fine-tuning the process. This culture change benefits both teams as well as the organization as a whole.

Figure 4 shows how insights from embedded analytics can be disseminated to transform a company into an effective learning organization.

Figure 4

 

Figure 4: An integrated framework for BI and advanced analytics supports best-of-breed business processes

Conclusion

Performance and insight optimization services from SAP can help you create an integrated analytics framework. Backed by deep industry expertise, our team of experts in data mining, mathematical modelling, business intelligence, and performance analytics can help design an analytics strategy and integrated analytics roadmap, blueprint, and target architecture. SAP experts can then help identify, design, and deploy the specific analytical applications that best support your business objectives. 

To report this post you need to login first.

Be the first to leave a comment

You must be Logged on to comment or reply to a post.

Leave a Reply