In today’s data-driven world, achieving effective business intelligence (BI) remains a significant challenge. As Jean-Jacques Rousseau poignantly observed, learning and utilizing information at the right time is crucial yet often elusive. For many IT managers, the frustration of inconsistent reports and data discrepancies is all too familiar. The root of the problem lies in the lack of data standardization, integrity, and timeliness, leading to subpar business intelligence.
This frustration is echoed by business managers who often express disbelief at the difficulties in generating accurate reports. According to Gartner, business professionals must make critical decisions based on vast amounts of internal and external data, often without the necessary analytical support. The result? Decisions are made in a virtual vacuum, relying on outdated or incomplete information.
As an IT manager, your challenge is to bridge the gap between the technical limitations and the business needs, helping your colleagues understand the complexities involved and the steps necessary to improve the situation.
Take Action Before It’s Too Late
Addressing this issue head-on is essential to prevent your credibility from being questioned. Start the process of improving information management immediately, even with limited resources. The first steps are relatively simple and manageable, laying the foundation for a more comprehensive BI strategy. The key is to begin now.
Steps to Improve Business Intelligence
Not all information management processes require these steps, but most organizations will benefit from implementing them:
- Raise Awareness of Barriers to Better Business Intelligence: Educate your team about the challenges of achieving high-quality BI.
- Educate Business Units on the Costs and Benefits of High-Quality Data: Help them understand the value of accurate and timely data.
- Establish Business Principles for Data Oversight: Develop guidelines for monitoring data quality.
- Justify the Need for Data Infrastructure: Rationalize the need for a streamlined data management system.
- Cleanse and Consolidate Databases: Improve data integrity through systematic cleansing.
- Define Organizational Responsibilities for Data Management: Assign clear roles for ongoing data maintenance.
- Extend Principles to Data Warehouse Management: Apply oversight principles to any data warehouses.
- Build a Data Warehouse with Proper Support: Ensure a well-supported infrastructure for data management.
These steps are explored in detail below. Avoid the temptation to jump straight to building a data warehouse; instead, focus on establishing a solid foundation first.
Raising Awareness of Barriers to Business Intelligence
When confronted with complaints about data inconsistency, redundancy, or distribution across multiple systems, take these concerns seriously. Avoid technical jargon that might confuse non-IT staff. Instead, use simple graphical representations to explain these issues, making the concepts accessible and understandable.
For instance, use graphics to demonstrate how data redundancy or inconsistency affects information management. These visual aids can be far more persuasive than lists of file names or systems, helping to build business support for your proposed solutions.
Consider publishing these explanations in collaboration with your IT support unit or through your organization’s newsletter. Engaging with the editorial team to include a dedicated IT section could further enhance understanding across the organization.
Explaining the Why and the How
It’s crucial to explain why improving business intelligence is necessary before delving into how it can be done. The primary reason is to enable business units to access accurate, timely information, facilitating better decision-making and day-to-day operations.
Quantify the costs of poor data management where possible. For example, a medium-sized organization might find that eliminating unnecessary data management costs could save substantial amounts annually. Demonstrating such savings to your CEO can underscore the importance of good data management.
Establishing Principles for Data Oversight
Once the need for better BI is understood, it’s time to introduce solutions. The first step is to agree on principles for managing data integrity and accuracy. These principles might address questions like:
- Should data be treated as a resource for the entire organization or just a specific department?
- What rules govern data archiving or deletion?
- Who is responsible for correcting identified data issues?
- Is there a logical repository for critical data?
These may seem like simple questions, but answering them can reveal deeper organizational issues. For example, if data is considered a corporate asset, who is responsible for maintaining its standards? Who evaluates data integrity, and who fixes errors in a flawed database? Understanding these dynamics is crucial for developing effective data management practices.
Rationalizing the Data Infrastructure
Simplifying the physical complexity and location of data should be a priority. While this is rarely a quick fix, developing strategies to consolidate and standardize data access can significantly improve data management. For example, creating centralized tables or files that multiple systems can access and update can reduce redundancy and improve consistency.
Data Cleansing
Data cleansing should be an ongoing process, guided by well-defined organizational data standards. Consider the following when planning your data cleansing efforts:
- Can you purchase error-free versions of your data from external sources?
- Should you use automated tools for data cleansing?
- Would it be more efficient to re-enter the data rather than sort through and correct it?
- Are there incentives for stakeholders to maintain clean data?
- Can data cleansing be integrated into regular transaction processes, focusing on data that doesn’t meet organizational standards?
Data cleansing should not be treated as a one-time task. Business units must understand how to manage data maintenance effectively to prevent future issues. For example, some organizations remove the responsibility for entering critical data from frontline operators and assign it to specialized staff with access to reference materials and editing tools.
Implementing a Data Warehouse
If you have a compelling reason to build a data warehouse, ensure that you do so carefully. Don’t rush into it until you’re confident that the data being fed into the warehouse is accurate and relevant. Data warehouses are often filled with irrelevant data, so focus on defining the attributes that truly matter to the business.
Design your data warehouse to link these attributes to relevant information, ensuring that each attribute is well-assessed. A key question to ask is, “How will this information contribute to the profitability of the business?” Expect to encounter surprising answers.
Avoid being misled by the low cost of disk space; the goal is to analyze and leverage data for business intelligence, not merely to store it. Ensure that your data warehouse design goes through rigorous checks to guarantee data quality.
Finally, clearly define the data warehouse management team’s roles, qualifications, responsibilities, and duties. This team should focus on managing data, not technology, and should work closely with their business counterparts.
Conclusion
One of the most frustrating aspects for your business colleagues is the lack of effective business intelligence. They are often forced to make critical decisions based on incomplete, inaccurate, or inaccessible data. They recognize the problem but can’t understand why it hasn’t been resolved.
By taking the initiative and showing them a clear path forward, you can begin to address these issues. Although the journey is challenging, the rewards are significant if done correctly.
Gartner highlights that data platforms, data cleansing, and system integration are among the weakest areas in business intelligence systems. Additionally, they emphasize that the four key quality metrics for information are relevance, timeliness, completeness, and consistency. With the growing importance of the web in business intelligence, it’s crucial to be prepared.
Checklist
- Have you established principles for information management?
- Are these principles communicated across the organization?
- Does the organization have standards for managing different types of data (e.g., titles and abbreviations)?
- Is there a logical justification for the number of data entry points?
- Can the number of files and systems be rationalized to maintain data integrity?
- Can error-free static data (e.g., names and addresses) be purchased from a vendor?
- Are data entry personnel motivated to ensure data accuracy?
- Is there a plan to penalize those who enter data carelessly?
- Will improving data collection tools enhance data quality?
- Can you implement a system where users select data from lists rather than typing it in?
- Have you established a data maintenance team responsible for maintaining high-quality data?
- Are the responsibilities for data maintenance clearly defined for the relevant department managers?