Using Open Source Platforms for Business Intelligence

Avoid Pitfalls and Maximize ROI

Lyndsay Wise

• Publisher: Morgan Kaufmann
• Year: 2012
• ISBN:
  9780124158764
• Topic: Business
• Subtopic: Business Intelligence
• Index: Business

Open Source BI solutions have many advantages over traditional proprietary software, from offering lower initial costs to more flexible support and integration options; but, until now, there has been no comprehensive guide to the complete offerings of the OS BI market. Writing for IT managers and business analysts without bias toward any BI suite, industry insider Lyndsay Wise covers the benefits and challenges of all available open source BI systems and tools, enabling readers to identify the solutions and technologies that best meet their business needs. Wise compares and contrasts types of OS BI and proprietary tools on the market, including Pentaho, Jaspersoft, RapidMiner, SpagoBI, BIRT, and many more. Real-world case studies and project templates clarify the steps involved in implementing open source BI, saving new users the time and trouble of developing their own solutions from scratch. For business managers who are hard pressed to indentify the best BI solutions and software for their companies, this book provides a practical guide to evaluating the ROI of open source versus traditional BI deployments.

• The only book to provide complete coverage of all open source BI systems and tools specifically for business managers, without bias toward any OS BI suite
• A practical, step-by-step guide to implementing OS BI solutions that maximize ROI
• Comprehensive coverage of all open source systems and tools, including architectures, data integration, support, optimization, data mining, data warehousing, and interoperability
• Case studies and project templates enable readers to evaluate the benefits and tradeoffs of all OS BI options without having to spend time developing their own solutions from scratch

SECTION 1

Introduction and market overview

Chapter 1 Introducing BI

Chapter 2 OS overview

Chapter 3 The convergence of OS and BI

Chapter 4 A look at the OSBI market

Chapter 1

Introducing BI

Why is there a need for this book?

Many books exist that identify how to get the most out of analytics or how to develop an open source business intelligence (OSBI) solution based on specific development or solution requirements. The reality is that even though these books provide value within the niche they address, very little exists that provides an understanding of OSBI at the business level including:

1. What OSBI is and how its adoption can benefit organizations;

2. The general considerations required before embarking on an OSBI project and the potential challenges that exist;

3. How to justify an OSBI project from the business side (instead of from an IT development perspective); and

4. The tools required at the business decision-making level to make sure that your project is a success.

In a 2011 study of 163 companies by Project Management Solutions,¹ it was estimated that 37% of general projects are at risk of failure. Within IT related projects specifically, this number is much larger. In a 2010/2011 interview study by Geneca,² they found that 75% of participants lacked the confidence to achieve project success from the start, with 80% of professionals admitting that much of their time is spent on project rework. In essence, organizations are embarking on projects they may not be ready to tackle. Project failures, reworks, missing timelines, and going over budget are all signs of mismanagement, a lack of processes and best practices, and a misunderstanding of the requirements and end goal. Unfortunately, business intelligence (BI) is no different. And adding open source (OS) to the mix doesn’t make things easier. In many cases, it is the opposite because of the promise of “free” and what that really means.

So, how will your organization ensure that you are in the 25% range of organizations confident and successful when implementing or expanding your OSBI environment? The reality is that there are no guarantees. But, being armed with information and an understanding of what tools you need to get there will help your business develop the proper approach to implementing an OSBI solution or, alternatively, deciding that it isn’t right for your business. Either way, knowledge is power, and this knowledge has not existed in a cohesive guide to help business decision makers make sense out of business intelligence, OS, its overlaps, and how to make the right decision for the organization as a whole. In the past, OS was a tool used by IT developers, but with the market slowly shifting towards self-service applications, business users require a greater understanding of how these solutions will benefit them by creating better visibility and higher productivity.

As an industry analyst and consultant working primarily with small and mid-sized businesses looking at implementing business intelligence solutions, I continue to see collaboration among c-level executives and IT directors when making technology and specifically, BI decisions. Because many IT developers like the promise of developing their own OS solutions for the business, many business decision makers hear the words “free software” and jump on the bandwagon. Without much else to go on, OSBI is selected. The hidden costs, extra time to develop, and other considerations seem to be outweighed by lack of initial software costs. Nowhere does the adage “free as in a puppy and not as in free beer” apply more directly than within a BI project. Business intelligence requires many separate components to make it work. Knowing how to put these together and understanding all of the areas where OS offerings may fit mean that deciding to implement OSBI requires more than a cursory understanding of the market.

The goal of this book, therefore, is to help you and others within your company make the right decisions when it comes to OS adoption. Is OSBI right for you? How is free software offset by development efforts and long-term maintenance? Is OSBI more strategic than traditional BI offerings? And how does OSBI differ from the broader business intelligence market landscape? All of these questions will be addressed while guiding you through the process of what it takes to successfully start and complete your BI initiatives. All of these questions so far are just the tip of the iceberg in terms of what needs to be asked when evaluating OS BI – or any BI project for that matter.

What to expect in this book

The overall goal of this book is to help you and your colleagues make the right BI-related software decisions. In order to do this, this book aims to address three key areas:

1. To provide an understanding of the OS market and OSBI specifically. This gives business decision makers a broader understanding of the OS market and its potential value to the business.

2. To identify what benefits, challenges, and efforts are required in OSBI development on a high level so that business decision makers will know what questions to ask.

3. To link the value proposition of BI with OS adoption – how to justify projects and make sure you have the resources you require to get the company from project initiation through BI development and delivery successfully.

Therefore, this book is broken into sections to provide easy access to your areas of interest. After reading through the sections it will be easy to go back to the areas that are of particular interest.

Section 1: Introduction and Market Overview provides an introduction of BI and why it cannot be overlooked. This section also provides an in-depth look at the OS market, BI OS, and the general options available. The importance of a general overview is not only to understand the market, but also to gain a broader perspective of how OSBI potentially meets the business requirements of your organization. In addition, because source code is provided for free, many solution providers tout that these are free solutions, when the reality is that development, maintenance, and additional hardware are costs that may be overlooked if decision makers do not understand the details involved within OS projects.

Section 2: A Deeper Look at Open Source BI goes a little deeper by providing a comparison of general OS and commercial OS offerings. This includes looking at the differences and the connotations of selecting one model over the other. This section also tackles the business benefits and challenges of OSBI and looks at why OSBI is becoming more of a contender within the realm of BI adoption. Because OS represents a niche area within the broader BI market, what it has to offer organizations is different from other business intelligence offerings. In addition, we will look at the increase in OS popularity and how it affects the BI market specifically and what it means for your organization should you choose to adopt OS.

Section 3: BI Strategies for Success – Tying in Open Source Adoption with BI Success discusses the types of companies that might benefit from OSBI adoption. Business decision makers require a broad understanding of the market to help identify how solutions compare with one another, whether OS or traditional solutions. This means defining the implications of adoption and use.

Section 4: Justifying OSBI Projects continues the theme from Section 3 by looking at how to sell an OSBI project to other business units by delving into the return on investment (ROI) and total cost of ownership (TCO) and by looking at ROI and TCO models that apply. This involves taking a step back and looking at what constitutes ROI and TCO within business intelligence projects and how they differ with OS specifically. This includes comparing options and looking at some of the differences in cost, internal resources required, and development and maintenance efforts. In essence, we take some of what we learned from Section 2 and transform it into practical next steps to evaluate potential OSBI offerings. Generally, ROI and TCO identification also enable stronger justification of project costs and initiation. Whether business or IT driven, project sponsors need to be able to justify their project costs and tie that to what they hope to achieve with their overall BI initiative. Justifying costs and project goals and tying that to increased visibility within the organization helps pave the way towards initial BI and expansions over time.

Section 5: Understanding the Technology Behind the Business Value takes a turn to identify some of the technical considerations in hopes of encouraging broader IT/business unit collaboration. Even though developers create the dashboards, analytics, and reporting, their main goal is to provide a strong data infrastructure. Because data integration, data management, and database design are the components that work towards building a strong BI infrastructure, they are impossible to overlook. Therefore, this section provides a high-level look at the general data integration requirements, what it means to work within an OS environment, what skillsets are required, potential challenges that exist, and some of the benefits of selecting OSBI. For business sponsors, this knowledge will increase the cohesion of IT/business unit relations and help business managers get a better understanding of all of the considerations and effort required to implement an OSBI offering.

Finally, Section 6: Takeaways/Recommendations puts all of the pieces together and provides some next steps and practical advice on how to get from software selection to implementation. Aside from the resources available through OS community participation, there are many resources that exist that businesses can take advantage of.

Now that we’ve looked at the purpose of this book and what it includes, let’s get started! The first step is to separate ourselves from the topic of OS and to take a step back and look at business intelligence, what it is and why you need it. Once we understand its value independently, we can look at how the OS market has shifted some of the available offerings and the general expectations of organizations.

An introduction to BI

Overall, the software industry provides a broad range of solutions to meet the needs of companies, ranging from transactional and supply chain management to customer relationship management (CRM) and project planning. Organizations apply a wide variety of software throughout their companies to manage their daily operations. Unfortunately, many of these solutions are implemented independently of one another and use different infrastructures that do not easily integrate with one another. This makes it hard to understand customer lifecycles, identify broader supply chain or sales opportunities, or create a single view of information to enable better planning. Even though individual departments can generally get the general insights they need to develop the high-level metrics they require, the ability to integrate sales information with supply chain requirements or the ability to look at marketing campaigns and customer sentiment or product successes by geography may be nonexistent. Add to this the adoption of social media and continual data volume growth and using these channels to improve branding and to tie in sales and marketing initiatives and businesses may be hard-pressed to create a holistic view of performance.

Executives and business decision makers in this situation may be flying blind. In the past, executives were able to make their decisions and plan initiatives based on comparing printed static reports and conversations with business managers. This is no longer the case. A fast-paced competitive marketplace where customers (both businesses and consumers) can purchase products and services from anywhere in the world means that organizations can no longer overlook the use of analytics. Businesses require a way to centralize information assets from across the organization as well as access data from external data sources, such as through partner channels, suppliers, or on social media sites. Consolidating this data from disparate data sources leads to gained insights that are needed in order to remain competitive within a constantly changing competitive landscape.

Enter business intelligence. Figure 1-1 will provide an overview of how BI enables organizations to consolidate data from various sources, manage both data quality and business processes, and develop a series of analytics through interactive, front-end applications. Overall, traditional BI solutions have various layers but provide a way for systems that reside within disparate data sources to be integrated into a single database. The goal of this is twofold: One, to create a source of data storage and analysis that is separate from transactional data sources, and two, to enable historical data capture while limiting storage requirements that affect operational systems. Both of these combined provide the basis for a BI infrastructure. Once this infrastructure is put in place, organizations can develop the analytical layer to identify trends, manage performance, look at sales analytics, or develop thresholds as a way to alert employees of potential issues proactively. And nowadays, BI extends further into the realm of operational intelligence, where businesses look at monitoring the business in real time to adopt a more proactive approach to managing key performance indicators (KPIs) and overall performance.

Figure 1-1 Traditional BI diagram – Identifying the design of BI solutions. © WiseAnalytics 2011.

Although considered by many as a software that supports operations, the reality is that much effort is required to develop and maintain a full BI infrastructure in-house. The diversity of solutions available include both traditional BI environments and targeted solutions that are developed to satisfy a specific need or to address a defined business problem. Niche applications also exist that include industry-specific offerings or corporate performance management. But some of the most common BI applications include sales and marketing analytics, call center monitoring, and budgeting and forecasting. For instance, let’s say a small retailer is trying to understand why sales for a particular product has dropped in 2 of their 24 stores but remained steady at the other locations. They might have to look at sales history to identify whether this is normal performance historically, whether circumstances have changed at those two locations, inventory, foot traffic, placement of items, and how these compare to the better performing locations. In order to gain this amount of visibility it becomes important to centralize varying sets of data. And what happens if weather played a factor? These two locations might be located in a dry and sunny region that has met with higher than average rainfall. In this instance, there might not be a need for customers to buy extra tanktops or sunglasses, providing a simple explanation for the drop in sales. Whether or not the answer is simple, the reality is that organizations require as much information as possible to better plan and manage their business.

In the beginning of all of this two disparate approaches to BI design became the de facto methods of development. On a high level, Bill Inmon,³ known as the father of data warehousing, believed in the concept of building a centralized dat...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Dedication
• Acknowledgments
• SECTION 1: Introduction and market overview
• SECTION 2: A deeper look at OSBI
• SECTION 3: BI strategies for success – tying in OS adoption with BI success
• SECTION 4: Justifying OSBI projects
• SECTION 5: Understanding the technology behind business value
• SECTION 6: Takeaways/recommendations
• Index