|
Page 1 of 2
Many enterprise decision makers are hesitant to even consider a BI initiative because of the anticipated cost. Here are four ways to implement BI technology without breaking the bank.
Also See:
Business Intelligence The Intelligent Way
What BI Tool Is For You?
This article was originally published by Info-Tech Research Group. Copyright (c) 1998-2008 Info-Tech Research Group. All rights reserved. Reprinted by permission.
Executive Summary
Contrary to popular belief, a BI initiative need not be painfully expensive. An enterprise can implement an effective BI solution consisting of existing software and reasonably-priced technology. This research note shows how to keep costs down by addressing the following areas:
- Cost-effective reporting and analysis tools.
- Affordable data mining and Online Analytical Processing (OLAP) solutions.
- Low-cost BI software suites.
- An economical approach to data warehousing.
In light of these low-priced alternatives, a BI initiative may not be beyond the enterprise's budget after all.
Strategy Point
A BI initiative is a serious undertaking, and often an expensive one. All enterprises can benefit from improved BI, but many are deterred from taking action because of the prohibitively high costs involved. IT and business decision makers need to understand that BI tools are not all expensive. It is entirely possible to implement an enterprise BI solution consisting of existing tools and affordable technology, particularly for organizations with a Microsoft-heavy software portfolio.
Key Considerations
As discussed in the ITA Premium research note, "Business Intelligence the Intelligent Way," most BI tools fit into one or more of the following functional categories:
- Reporting and analysis software.
- Data mining and Online Analytical Processing (OLAP) tools.
- Comprehensive BI suites.
- Data warehousing solutions.
The four tables below present the traditional technology strategy for each category contrasted with a more cost-effective option.
Table 1. Cost-Effective Reporting and Analysis
| Traditional Approach |
Cost-Effective Alternatives |
| Dedicated tools for basic reporting, analysis, and dashboarding. When enterprises consciously take on a BI initiative, they typically turn to software vendors that specialize in BI tools. If compliance requirements are not strict and analytic needs are not sophisticated, a simpler solution may suffice. |
Use existing Microsoft Excel features. With MS Office being the world's dominant productivity suite, most enterprises users already have Excel and know how to use it. Enterprises that opt for Excel as a BI tool will be able to minimize license costs and reduce training costs as well, although users will require training on BI-specific usage. Out of the box, Excel 97 to 2007 can be employed for the following:
Reporting and analysis. The pivot table function in Excel facilitates aggregation and summarization of data. Although it is not as robust as most dedicated BI tools, Excel can produce effective reports and enable multidimensional analysis. Excel 2007 can maintain a persistent connection with the data source, ensuring currency of information.
Dashboarding. At its core, Excel is a data visualization tool. A dashboard is a visual summary of multi-source performance data. Excel is fully capable of producing executive-friendly dashboards, but the creation process is not intuitive. ExcelUser offers training materials on Excel dashboards priced at under $70.
Enterprises that have MS Access can use it as an ad hoc query and reporting tool instead of or in addition to Excel. Unlike Excel, Access allows ETL activities using relational database techniques. Access can also run against SQL so that the user does not have to import if tables are too large.
Enterprises that concurrently run MS Office SharePoint Server 2007 Enterprise Edition will have the added benefit of Excel Services, a tool designed to extend Excel's BI functionality. One of the greatest advantages of Excel Services is that it enables widespread sharing of workbooks, reducing the proliferation of unsanctioned data.
Enterprises that run SQL Server should use the Reporting Services component, which is comparable to dedicated BI reporting software and much more robust than Excel alone.
|
| Price range: About $500 per seat. |
Price range: Excel is included with all editions of MS Office. Access is included with the Professional and Ultimate editions. Each application is $229 per seat if purchased separately.
|
| Source: Info-Tech Research Group |
Table 2. Cost-Effective Data Mining and OLAP
| Traditional Approach |
Cost-Effective Alternatives |
| Dedicated OLAP and data mining tools. When it comes to multidimensional analysis and data mining, enterprise decision makers often believe specialized tools are superior to database server software with added BI functionality. From a pragmatic perspective, it does not make sense to adopt dedicated tools if an existing solution can meet the enterprise's analytic needs. |
Use the existing SQL Server implementation. According to Microsoft, about 75% of enterprises run SQL Server databases, and much of the recent growth in market share has been driven by the functional improvements in the latest version. The Analysis Services component includes two major BI functions:
OLAP. Analysis Services began in 1998 as OLAP Services in SQL Server 7. It currently supports all three storage modes: multidimensional OLAP (MOLAP), relational OLAP (ROLAP), and hybrid OLAP (HOLAP). The OLAP tool includes built-in wizards to facilitate cube creation. Queries are typically performed using Multidimensional Expressions (MDX), but an Excel Pivot Table add-in is also available.
Data mining. OLAP Services became Analysis Services with the release of SQL Server 2000, when Microsoft added data mining capabilities. SQL Server Data Mining is a platform for building and running extensible custom mining models. Refer to the MSDN site for the set of data mining algorithms available in the current version.
|
| Price range: About $800 per named user (including both data mining and OLAP). |
Price range: Likely already installed (Standard edition is $6,000 per processor).
|
| Source: Info-Tech Research Group |
Table 3. Cost-Effective BI Suites
| Traditional Approach |
Cost-Effective Alternatives |
| Commercial BI suites. Off-the-shelf packages such as the offerings from Oracle, IBM Cognos, and SAP are comprehensive and powerful, but they are also expensive. Even midsize-friendly options like Microsoft PerformancePoint and QlikView are pricy compared to the open source alternatives. |
Use open source software. Open source BI solutions have matured to the point where they will meet the reporting and analysis needs of many enterprises. They are also extensible and interoperable with third-party tools, so functional limitations are no longer deal-breakers. There are two offerings that are competitive as BI suites:
Pentaho is a well-established commercial open source project with a number of large clients in the private and public sectors. Major functional areas include reporting, dashboarding, OLAP, data mining, and data integration. Pentaho provides a range of commercial services, including software support, training, and consulting. System integration services are available from Pentaho partners worldwide.
JasperSoft claims the greatest market share of all open source BI offerings. Clients span a wide range of industries around the world. Functionality includes reporting, dashboarding, OLAP, and data integration. The JasperSoft BI Suite is available in Community (unsupported) and Professional editions. The latter includes added features as well as support plans. JasperSoft also provides training and consulting as commercial services.
|
| Price range: Roughly $100,000 to $500,000 for the enterprise (varies greatly depending on number of users and role distribution). |
Price range: No license cost; optional support subscription for about $10,000 to $15,000 per year.
|
| Source: Info-Tech Research Group |
Table 4. Cost-Effective Data Warehousing
| Traditional Approach |
Cost-Effective Alternatives |
| Customized data warehouse. Arguably the single greatest deterrent for large-scale BI initiatives is the astronomical cost of data warehousing. License costs for the software alone are daunting. Infrastructure costs are usually even higher. Enterprises that must centralize their data in a data warehouse now have more affordable options. |
Buy a midmarket data warehouse appliance. With the advent of these appliances, enterprises can now build and maintain a fully functional data warehouse for a fraction of the cost. Strictly speaking, the Teradata Enterprise Data Warehouse can be regarded as an appliance, but it is a high-end solution with a high-end price tag. The following vendors offer a more reasonable pricing structure:
Netezza is the current midmarket leader, known for its ability to handle high data volumes at low cost. Netezza's uniqueness lies in its streaming analytic technology, which allows complex non-SQL algorithms to be processed within the data warehouse.
Sun Microsystems is a strong competitor with a data warehouse appliance that uses open source Greenplum Database software. Sun boasts the highest storage density among appliances, at 24 TB in 4 U.
Dataupia is a relatively new entrant to the data warehouse market. Dataupia is designed for persistent data access, which enables it to consistently outperform competing appliances when managing continuously changing data.
|
| Price range: Varies greatly depending on size of implementation; often in the millions. |
Price range: $20,000 to $50,000 per TB.
|
| Source: Info-Tech Research Group |
Next: 3 Recommendations
|
